1

Projektrapport från utbildningen i

EKOTOXIKOLOGI

Ekotoxikologiska avdelningen

Nr 109

Risk assessment on butylphenol, octylphenol and nonylphenol,

and estimated human exposure of alkylphenols from Swedish fish

Beatrice Jonsson

Preface This report is a graduate project in ecotoxicology from the Department of Environmental Toxicology, Evolutionary Centre, Uppsala University, performed at the Division of Toxicology at the National Food Administration (NFA). I would like to thank all the staff at the Division of Toxicology, National Food Administration for making my staying there most pleasant. A special thanks to my supervisors Anders Glynn and Per-Ola Darnerud for their guidance and valuable comments during this project. I would also like to thank Nils Jansson at the NFA library for the help in finding articles. Finally, I would like to thank Jan Örberg at the Department of Environmental Toxicology, Uppsala University for the final reading of this project.

2

Abstract Present report is a review of the knowledge about the alkylphenols butylphenol (BP), octylphenol (OP) and nonylphenol (NP) with focus on mammalian toxicology and human exposure. The report includes an estimation of human OP and NP intake from Swedish fish. Because of few available data on BP, no conclusion could be drawn about human health risks due to BP-exposure. Alkylphenols have low acute toxicity, OP and NP are not genotoxic and have low carcinogenic potential. OP and NP have been found to be estrogenic in several in vitro and in vivo systems, where OP is most potent of the alkylphenols. The critical effects due to OP exposure are changes in sperm morphology in rats (LOAEL 20 ng/kg bw/day) and increased length of gestation in pigs (LOAEL 0.01 mg/kg bw/day). The critical effect of NP exposure is increased kidney weight (LOAEL 15 mg/kg bw/day). In present study, tolerable daily intake (TDI) of OP is suggested to 0.067 ng/kg bw/day and 33.3 ng/kg bw/day for men and women, respectively. A TDI for NP is suggested to 50 000 ng/kg bw/day. The general population is mainly exposed to OP and NP from food, especially fish, and to some degree drinking water. For NP, migration from food wrapping plastics may also be an important exposure source. In Swedish freshwater fish, median concentration of OP was 0.15 ng/g f.wt in perch, 1.2 ng/g f.wt in artic char and 11 ng/g f.wt in bream from a polluted area. For NP, the median concentration was 0.99 ng/g f.wt in perch and 15 ng/g f.wt in perch from a polluted area. A realistic estimation of average OP intake from fish for the Swedish population is within the range 0.011-0.1 ng/kg bw/day. A mean worst-case intake of OP and NP from Swedish fish is 3.66 and 5.23 ng/kg bw/day, respectively. Based on animal data, the safety margin between intake of OP from fish among Swedish men and exposure levels causing effects on sperm quality in rats is low. However, the data on OP effects on sperm quality in rats are uncertain and the single study on rats needs to be replicated before firm conclusions can be drawn about sperm effects. In rat studies, no adverse effect on reproduction ability has been reported. The risk for negative effects on sperm quality in men due to OP exposure is unknown. Women have higher average OP exposure from fish than men, but are not at risk of exceeding TDI suggested in present study. A MOS (margin of safety) between case intake of OP from fish (45.4 ng/kg bw/day) and LOAEL for increased length of gestation in pigs is about 200. The risk for human effects due to exposure to NP is low, where a MOS between worst-case intake of NP from fish and LOAEL for kidney weight in rats of 15 mg/kg bw /day is about 230 000-fold.

3

Table of contents Preface .................................................................................................................. 2 Abstract ................................................................................................................ 3 Abbreviations....................................................................................................... 6 Background.......................................................................................................... 7

Physical and chemical properties............................................................................................ 7

Uses and sources of environmental contamination ............................................................... 8

Environmental occurrence and fate of alkylphenols ............................................................ 9 Water and sediments .............................................................................................................. 9 Sludge ................................................................................................................................... 10 Soil........................................................................................................................................ 10 Air......................................................................................................................................... 10 Bioconcentration .................................................................................................................. 11

Levels in food .......................................................................................................................... 12

Human intake ......................................................................................................................... 14

Kinetics and metabolism of alkylphenols............................................................................. 16 Uptake .................................................................................................................................. 16 Distribution and excretion ................................................................................................... 16

Toxic effects on laboratory animals and in vitro systems................................................... 19 Acute toxicity ........................................................................................................................ 19 In vitro genotoxicity studies ................................................................................................. 19 In vivo genotoxicity studies .................................................................................................. 20 Subchronic exposure ............................................................................................................ 20 Chronic exposure and carcinogenic effects ......................................................................... 20 Reproductive and developmental effects .............................................................................. 21

Environmental toxicity .......................................................................................................... 25

Human toxicity studies........................................................................................................... 26

Risk assessments and regulations ......................................................................................... 26 Present study...................................................................................................... 27

Aim of study............................................................................................................................ 27

Materials and methods........................................................................................................... 27 Fish samples ......................................................................................................................... 27 Calculations and statistics ................................................................................................... 28

4

Results ..................................................................................................................................... 31

Derivation of TDI ................................................................................................................... 36

Discussion................................................................................................................................ 39 References .......................................................................................................... 43

5

Abbreviations AP Alkylphenol(s) APE Alkylphenol ethoxylate(s) BCF Bioconcentration factor BP Butylphenol, 4-t-BP bw body weight d.wt dry weight EC50 median effective concentration f.wt fresh weight LC median lethal concentration 50 LOAEL Lowest Observed Adverse Effect Level LOEC Lowest Observed Effect Concentration MOS Margin Of Safety NFA National Food Administration NOAEL No Observed Adverse Effect Level NOEC No Observed Effect Concentration NP Nonylphenol/ 4-NP, 4-t-NP NPE Nonylphenol ethoxylate(s)/, NPE1 nonylphenol monoethoxylate OP Octylphenol/ 4-t-OP, 4-OP, t-OP, p-t-OP TDI Tolerable Daily Intake

6

Background This study focuses on the toxicology of butyl-, octyl-, and nonylphenol. Toxicological behaviour of other alkylphenols may to some extent be predicted by structural and physiochemical similarities with the described alkylphenols in the text. However, data shows differences in toxicological properties between isomeric forms of alkylphenols, which in some cases may make predictions about toxicity difficult to use. Alkylphenols are used as alkylphenol ethoxylate derivates (of which 80% consist of nonylphenol ethoxylates) in detergents, cleaning products and emulsifiers [1]. In water treatment plants alkylphenol ethoxylates are degraded to (non ethoxylated) alkylphenols, which are found in the environment [2]. Alkylphenols, mainly nonylphenol, have been found in high concentrations in water and sediments [3]. Adsorbed alkylphenols in sediments are slowly degraded. Alkylphenols are highly toxic to water organisms and have been observed to bioaccumulate in fish, crustaceans and algae [4]. Alkylphenols may cause endocrine effects, and may affect reproductive ability of aquatic species, and also in mammals [5, 6]. Nonylphenol and octylphenol have been discovered in drinking water and certain food products [7, 8]. Bioaccumulation of alkylphenols in aquatic species has raised concern for human health regarding intake of alkylphenols, especially from fish. However, the potential health effects in man are largely unknown.

Physical and chemical properties The basic structure of alkylphenols (APs) is a phenol ring with a hydroxyl group, and an alkyl chain. The alkyl chain varies in length, where the most common alkylphenols are butyl-, octyl- and nonylphenol with four, eight and ten carbon atoms (see Fig. 1.). The dominant group of alkylphenols is the para- (or 4)-isomer [7]. This means that the hydroxyl group is situated on the phenol ring opposite the alkyl chain. The term tert (t) refers to compounds with branched alkyl chain. The structure of alkylphenol ethoxylates (APEs) is a fat soluble phenol ring with a varying alkyl chain, and a water soluble chain of 1-100 ethoxylate groups (1EO-100EO) [7]. The mostly used APEs, nonylphenol ethoxylates (NPEs) has an alkyl chain (C H9 19, usually branched) and a chain of 9-10 ethoxylate groups (see Fig.1) [9]. With decreasing length of the ethoxylate chain, bioaccumulation and toxicity of APEs increases [10]. Butylphenol (BP) (CAS number 98-54-4) has several different synonyms; 4-tert-butylphenol (4-t-BP), p-tert-butylphenol, butylphen, 1-hydroxy-4-tert-butylbenzene, 4-hydroxy-1-tert-butylbenzene, 2-(p-hydroxyphenyl)-2-metylpropane and 4-(1,1-dimethylethyl)-phenol [9, 11]. BP has an empirical formula of C10H14O and a molecular weight of 150.2 g/mol. At room temperature BP is a white crystalline product. Melting point is around 98-101 ºC and boiling point around 236-238 ºC. Vapour pressure is 0.3 hPa at 50 ºC, density 0.91 g/cm3, water solubility 1.8 g/l at 25 ºC and log Kow is 3.29 at 25 ºC [11, 12]. Octylphenol (OP) (CAS number 140-66-9) is also named as 4-tert-octylphenol (4-t-OP), p-octylphenol, p-tert-octylphenol and p-(1,1,3,3-tetramethyl-butyl)-phenol. Empirical formula of OP is C14H22O and molecular weight 206.3 g/mol. Melting point of OP is 81-83 ºC and boiling point 280-302 ºC. Vapour pressure is 100 hPa at 20 ºC, density 0.92 g/cm3, water solubility 0.01 g/l at 25 ºC and log Kow is 3.7 [9].

7

Nonylphenol (NP) is a general name of several isomeric compounds, where isomers are named depending on position of the nonyl group on the phenol ring, and the degree of branching of the nonyl group. Technical NP is a mixture of different isomers, which also contains minor amount of octyl- and decylphenols [7]. NP is found under three different CAS numbers; 104-40-5, 25154-52-3 and 84852-15-3 (branched NP). NP is also named 4-nonylphenol (4-NP) or p-nonylphenol (p-NP) [1]. NP has an empirical formula of C15H24O and a molecular weight 220.34 g/mol. At room temperature NP is a light yellow liquid with a mild odour of phenol. Melting point of NP is 2 ºC, boiling point 290-302 ºC [9]. Water solubility and vapour pressure of NP are low, about 6 mg/l (20 ºC) and 0.03 Pa at 25 ºC, respectively [1, 9]. Density of NP is 0.95 g/m3 and log Kow about 3.28 (mean value for different reported values) [1].

O

CH2

CH2

OCH2

CH2

OHn

R

CH3OH

CH3

CH3

CH3CH3

CH3

CH3CH3

OH OH C9H19

A) B) C)

D)

Fig. 1. Chemical structures of some alkylphenols. A) 4-tert-butylphenol. B) 4-tert-ocylphenol. C) 4-nonylphenol. D) Alkylphenol (poly)ethoxylate (APE). R is an alkyl chain (usually branched) and n is the number of ethoxylate units. Nonylphenol ethoxylates (NPEs) has an alkyl chain (R) of C9H19 and a varying ethoxylate chain (n) of 1-100, where the mostly used NPEs have 9-10 ethoxylate units.

Uses and sources of environmental contamination APs and APEs have been used since 1950s. In United States, 11 000 tonnes BP were used in 1993 [12]. The net use of OP in European Union increased from 18 000 tonnes in 1997 to 23 000 tonnes 2001 [13]. For NP, the net use in European Union during 1994-1997 was fairly constant, 73 500 tonnes/year. During the same period, the NP-production within European Union decreased from 77 505 tonnes/year 1994 to 73 500 tonnes/year 1997. In United States, about 110 000 tonnes NP were produced or imported 2000 [9]. In Sweden 2003, the total use of 4-t-BP, 4-t-OP and NP were 6, 7 and 31 tonnes, respectively [14]. The main use of BP is in manufacturing of phenolic resins, as binders in manufacture of varnishes. Other uses are as antioxidant in motor oil and synthetic lubricants, emulsion breaker for petroleum oil, as a plasticizer for cellulose acetate in some plastic materials, and as an intermediate in chemical synthesis [15]. APs with short alkyl chain (such as BP) are polymerised by reaction with formaldehyde and used in plastics [9]. OP are used as an intermediate for production of resins, non-ionic surfactants and rubber additives, and in manufacturing of antioxidants, fuel oil stabilizers, adhesives, dyes, fungicides and bactericides [16]. NP is used in production of resins, plastics and stabilizers in industry [1], and has earlier also been used as softener in PVC-plastic in food packing materials [7, 17].

8

APs are mostly used in production of APEs in industries. Ethoxylation of NP starts by heating an alkali catalyst and evaporation of produced water (at about 120 ºC) [1]. Ethylene oxide is added under vacuum and reacts with all free NP. The length of the ethoxylate chain is varied by different ratio of NP and ethylene oxide or by controlling the reaction time. The use of ethoxylated APs (NPEs) are mainly as detergents in household products and in cleaners of electrical components and hard surfaces in industries [1, 8]. Other areas of use are as wetting agent in paper and pulp production, emulsifiers in paints and pesticide formulations, and in textile manufacturing for scouring, fibre lubrication and dye dwelling. Octylphenol-9EO (octoxynol- 9) is used in skin and hair products and as a spermicidal agent, and may be metabolised to NP. APEs are degraded to APs, a process that is a major source of environmental contamination. NPEs and NP in consumer products end up in wastewater and wastewater treatment plants, where NPEs are deethoxylated, mainly to nonylphenol mono- and diethoxylates (NPE1, NPE2) and NP [2]. During wastewater treatment, the reduction of NP and NPEs concentrations in wastewater is about 35-40%, where NP and NPEs are mainly distributed to sludge and effluents [2]. The total emission of NP and short-chained NPEs in Sweden after sewage water treatment has been determined to approximately 100-1000 tonnes/year [18]. Lower consumer and industrial use of BP and OP than of NP and NPEs results in lower concentrations in water treatment plants (typically more than 100 times lower), and therefore lower BP and OP concentrations in the environment. Exclusion and phase-out of NP and NPEs in Europe have been driven by voluntary agreements. Since 2003 there is a restriction to use of NP and NPEs in the European Union [19]. However, in the United States there is no restriction to the use of NP, which is believed to increase [9]. No reference regarding restrictions in the production or use of BP or OP has been found.

Environmental occurrence and fate of alkylphenols Some environmental levels of alkylphenols are presented in Table 1.

Water and sediments APs may reach the water recipient from water treatment plants and manufacturing plants, released in wastewater. Measured concentrations of OP and BP in river water and sediments suggest an accumulation in sediments, with more than 100 times higher concentrations of APs in sediments than in river water [3]. Based on physical and chemical properties, calculated half-life for BP and OP are the same, about 40 days in water and about 150 days in sediments [20]. Para-APs such as NP may end up in wastewater from their use in pesticide formulations, plastic leakage and from degradation of APEs [21]. Because of the hydrophobic nature of NP, persistence in the water column is low (half-life 1.2 days) due to distribution to macrophytes and sediment [22]. Persistence in sediments and macrophytes are higher with dissipation half-life of 28-104 and 8-13 days, respectively [22]. Hydrolysis and photolysis of NP in water are negligible [1]. NP is considered biodegradable, with an estimated degradation half-life in surface water of 150 days [1]. Biodegradation of NP in water and sediments have been studied by Eklund and co-workers [23]. Degradation rate in seawater increased after a month (0.06 to 1%/day), suggesting that

9

microorganisms need an adaptation period. In sediments, degradation rate was faster (1.2%/day) as a result of higher concentration of microorganisms. No information on degradation products of APs has been found. Degradation data of different NP isomers, suggests a slower degradation of more branched isomers [1].

Sludge Decrease of all NP compounds during sewage treatment is estimated to be 35-40% [2].Of the NP compounds, a majority of NPEs are discharged into the environment from sewage treatment plants via secondary effluents, and about 90 % of NP via sewage sludge [2]. The biodegradation increases with increasing temperature. Aerobic degradation is more efficient than anaerobic degradation [24]. Non-ionic and water soluble APs will partition between sludge and wastewater depending on Kow [9]. Aerobic degradation of NP and NPE1 has been studied in sewage sludge [25]. Degradation rate of NP was higher than that of NPEs. Gram-positive and rod-shaped bacterial strains, degraded NP and Bacillus sphacericus was most effective. Addition of yeast cells increased degradation rate. The degradation rate was slower at higher concentration of NP, where higher temperature resulted in faster and more complete degradation. NP was degraded (measured as percent remaining NP) within 21 days in untreated sewage sludge (pH about 6.4, temperature 30 ºC).

Soil With a high Kow, APs have high potential for retention in soil. For OP released to soil by application of sewage sludge, 99.6% of the OP concentration will remain in the soil and not be washed out [16]. Based on physical and chemical data, calculated half-life in soil is about 40 days for 4-t-BP and 4-t-OP [20]. Degradation half-live of NP in soil has been estimated to 300 days [1]. Soil adsorption and desorption of the APs 4-methylphenol, 2,5-dimethylphenol, 3,4-dimethylphenol and 2-tert-butylphenol were studied by Crespin and co-workers [26]. In neutral soils (pH 7.91) with high clay mineral content, APs were uncharged (non polar) and interacted strongly with clay particles. The recovery percent of applied AP concentration in these soils were low, about 4 %, suggesting a high AP adsorption by the soil. With higher and lower pH, increased soil contamination and a higher sand content, desorption of APs increased.

Air BP released to air was distributed to air (39.7%), soil (35.9%) and water (23.3%) [12]. Half-life of BP in air (hydroxyl radical degradation) has been reported to 0.4 days. Measured air concentrations of BP from an industrial area in Sweden ranged between 0.11 and 0.34 ng/m3

[20]. OP released to air is mainly distributed to soil (68%) and to air (26%) [16]. OP concentration in Swedish air is below the detection limit (0.3 ng/m3) [20]. Based on chemical and physical data, calculated half-life of 4-t-OP in air is about 0.26 days. With a possibly degradation with hydroxyl radicals, a calculated half live of NP in air is 0.3 days [1]. In New York, air concentrations of NP have been measured up to 69 ng/m3 [27]. In Sweden, NP has been measured in two air samples to 0.29 and 2.8 ng/m3, from an industrial and an urban area, respectively [20].

10

Table 1. Environmental levels of BP, OP and NP. STP= municipal sewage water treatment plant Medium Substance Concentration Reference Surface water µg/l Coastal water, Singapore (28 locations) 4-t-BP 0.01-2.3 [28] River water (US) 4-t-OP 0.2-2.0 [29]

Coastal water, Singapore (28 locations) 4-t-OP 0.01-0.80 [28] River Aire (UK) NP (total extractable) <1.60-180 [30] Receiving water from textile industry Coastal water, Singapore (28 locations) 4-NP 0.20-2.76 [28] Wastewater µg/l US chemical plant t-BP 1.00-150.0 [3] US chemical plant OP (2 isomers) 1.00-75.0 [3] Treated wastewater Sweden 4-t-OP 0.005-0.22 [31] Treated wastewater Sweden 4-NP 0.029-5.50 [31] Sludge mg/kg d.wt Switzerland STP NP 6-52.1 [32]

Sweden STP 4-t-BP <0.001-0.029 [20] Sweden STP 4-t-OP 0.076-8.7 [20] Sweden STP NP 1.7-437 [20] Sediment mg/kg d.wt Downstream US chemical plant 4-t-BP 0.2-7 (mg/kg) [3] Downstream US chemical plant, (one sample) OP (2 isomers) 5 (mg/kg) [3] Tees estuary, UK OP 0.017-0.34 [33] Germany, 23 locations NP 0.056-14.8 [34] Sweden 4-t-BP 0.0015-0.028 [20] Sweden 4-t-OP 0.00017-0.088 [20] Sweden 4-NP <0.006- 7.7 [20] Soil µg/kg d.wt Sweden (one site, 3 samples) 4-t-BP 0.6-1.7 [20] Sweden (one site, 3 samples) 4-t-OP 0.8-2.1 [20] Sweden (one site, 3 samples) 4-NP 11.0-61.0 [20] Air ng/m3 Coastal site (US) t-OP <1 [35] New York-New Jersey NP 0.2-69 [27] Sweden 4-t-BP 0.11-0.34 [20] Sweden 4-t-OP < 0.3 [20] Sweden (two samples) NP 0.29 and 2.85 [20]

Bioconcentration APs have a high potential of bioaccumulation in fish and mussels, with bioconcentration factors (BCF) above 100. Bioaccumulation increases with increasing Kow [36], and NP is bioconcentrated to a lager degree than BP and OP. Bioconcentration factor of 4-t-BP in carp (Leuciscus idus melanotus) and in algae (Chlorella fusca) have been estimated to 120 and 34 respectively [12]. BP has a BCF of 120 in fish [37].

11

Bioconcentration of 4-t-OP in rainbow trout (Oncorhynchus mykiss) after 10 days exposure in 4.0 μg/l during flow trough conditions, increased rapidly and reached a steady state within four days [21]. Major accumulating tissues were muscle (13%), bone (7%) and skin (5%). BCF (10 days, f.wt) was 800-1200 for fat, intestine, liver and pyloric caeca. In other parts of the fish BCF was below 300. In Atlantic salmon (Salmon salar) BCF (4 days) for p-NP was 280 [36]. Mean BCF (f.wt) in fatheaded minnow (Pimephales promelas) for five exposure concentrations (9.3, 19.2, 38.1, 77.5, 193 μg/l) was determined to 586±273 and 741±206 for 14 and 28 days, respectively [38]. BCF was found to be independent of concentration after 28 days exposure, but not after 14 days. Higher BCF of NP have been reported by Eklund and co-workers [39]. 14C-labelled 4-NP (solved in acetone) was added to water (flow trough conditions) at a concentration of 20 mg/l during 16 days, followed by an elimination period of 32 days. Samples were taken day 2, 4, 8, 16, 18, 20, 24, 32, and 48 from the start of the study. Common shrimp (Crangon cragon L.), common mussels (Mytilus edulis), and three-spined stickleback (Gasterosteus aculeatus L.) had BCFs (f.wt) of 110, 3400 and 1300, respectively. In the water plants Cladophora glomerata, Fontinalis antipyretica and Potamogeton crispus bioconcentration of NP is high, with BCFs ranging between 1000 and 10 000 [40]. Lower concentrations of NP found in fish compared to algae suggest that no biomagnification take place. NP was found in all tissues of one male wild duck (Anas boscas), where highest concentration was found in muscle (1.2 mg/kg d.wt).

Levels in food The main sources of AP exposure for the general population seem to be contaminated food, primarily fish, and to some degree also drinking water. Summary of measured concentrations of OP and NP in food are presented in Table 2. OP and NP have been measured in the molluscs, fish and crustaceans from different locations along the Italian coast [41, 42]. Edible parts of squid (Loligo vulgaris), cuttlefish (Sepia officinalis), mussels (Mytilus galloprovincialis) and clams (Chamelea gallina) contained NP concentrations of 67-696 ng/g f.wt [41]. Concentrations of OP were about 30 times lower, ranging between 2.7 and 8.6 ng/g f.wt. Highest AP contamination was found in squids, where average concentrations were 513 and 11.3 ng/g f.wt for NP and OP, respectively. In different fish and crustacean species, OP concentration ranged between 0.8 and 3.8 ng/g f.wt [42]. For NP, concentrations ranged between 12 and 1285 ng/g f.wt. Highest concentrations of NP were found in the fishes anchovy (Engraulis enchrascicos) and red mullet (Mullus barbatus). In a German study estimating daily NP intake from general food, the highest NP levels were found in apples (19.4 ng/g f.wt), tomatoes (18.5 ng/g f.wt) and butter (14.4 ng/g f.wt) [8]. NP was also found in breast milk (0.3 ng/g f.wt). The high concentration of NP in vegetables was suggested to depend on NP-containing pesticides. No correlation between fat content and NP concentration in food was found. NP and OP have been detected in fresh fruit and vegetables in Taiwan [43]. Highest NP-concentrations were found in plums, pears and apples with 5.9, 7.6 and 16 ng/g f.wt, respectively. 4-t-OP was found in broccoli (0.4 ng/g) and pear (0.7 ng/g). In other fruits and vegetables, concentrations of OP were below detection limit (0.2 ng/g f.wt).

12

APs have been found in seafood in Singapore [28]. 4-t-BP was found in the molluscs prawn (Penaeus monodon), crab (Portunus pelagius), blood cockle (Anadara granosa), white clam (Meretrix meretrix) and squid (Loligo sp.), and in fish (Decapterus russelli) bought from a local supermarket, in concentrations of 6.5-24.0 ng/g f.wt. Concentrations of 4-t-OP ranged between 6.7 and 44.9 ng/g f.wt, the highest concentration found in white clams. Highest 4-NP concentration was found in prawns and crabs (197.0 and 103.1 ng/g f.wt, respectively). NP concentrations in other species ranged between 46.6 and 64.8 ng/g f.wt. Several measured concentrations of NP in fish have been reported. In river water in Michigan (US) the highest concentration of NP in fish from contaminated waters was found in rock bass (Ambloplites rupestris), 8.1 ng/g f.wt [44]. Other fishes, such as iongnose sucker (Maxostoma macrolepidotum) and green sunfish (Lepomis macrochirus), had no detectable concentrations (detection limit 3.3 ng/g). In river water in Ohio (US) from a “clean” sample site upstream a wastewater treatment plant, NP concentration in carp (Cyprinus carpio) was of 6.6 ng/g f.wt [45]. NP concentrations in carp from all other sampling sites was 27 ng/g f.wt or above, where the highest NP concentration of 110 ng/g f.wt was found downstream a wastewater treatment plant. In a polluted area in Switzerland measured concentrations of NP in edible parts of fish ranged between 150 and 780 ng/g d.wt for Squalius cephalus, Barbus barbus and Salmo gairdneri [40], where estimated whole fish concentration for S. cephalus have been calculated to 35 ng/g f.wt [45].

13

Table 2. Measured mean concentrations of OP and NP in different types of food in ng/g f.wt. Seafood concentrations (ref. 41 and 42) are presented as range of mean concentration from three different areas along the Italian coast. Prawn, crab and white clam concentrations (ref. 28) are measured in samples bought from a local supermarket in Singapore. Food type Concentration OP Concentration NP Ref. /Species (ng/g f.wt) (ng/g f.wt) Seafood (edible parts) Mussels (Mytilus galloprovincialis), 4.4 and 4.9 254 and 265 [41] (Mean, two areas in Italy) Clams, (Ruditapes decussatus and 2.7-2.8 243-252 [41] Chamelea gallina) Squids (Loligo vulgaris) 3.9-18.6 389-696 [41] Cuttlefish (Sepia officinalis) 3.6-3.8 67-566 [41] Spottail mantis shrimp (Squilla mantis) 3.2-3.4 118-254 [42] Norway lobster (Nephrops norvegicus) 3.6-4.7 274-399 [42] Anchovy (Engraulis enchrascicolus) 0.8-1.7 497-1285 [42] Atlantic mackerel (Scomber scombrus) 2.6-3.8 1.162-270 [42] Red mullet (Mullus barbatus) 1.4-2.3 615-875 [42] Common sole (Solea vulgaris) 1.2-1.7 12-101 [42]

1.Prawn (Penaeus monodon) 20.4 (4-t-OP) 197.0 [28] Crab (Portunus pelagicus)1. 20.2 (4-t-OP) 103.1 [28]

1.White clam (Meretrix meretrix) 44.9 (4-t-OP) 46.6 [28] Other types of food Butter 14.4 [8] Whole milk 1.1 [8] Cream (30 % fat) 8.1 [8] Fresh cheese 7.5 [8] Milk chocolate 14.1 [8] Gooseberry marmalade 7.3 [8] Liver sausage 13 [8] Canned tuna 8.1 [8] Tomatoes 5.9 and 18.5 [8, 43] Apples 5.9 and 19.4 [8, 43] Plum 16 [43] Pear 0.7 7.6 [43] Broccoli 0.4 4.8 [43] Potatoes 0.6 [8] Pasta 1.0 [8] 1. Measured concentrations of 4-t-BP were also found: prawns 21.3 ng/g f.wt, white clams 6.5 ng/g f.wt and crabs 24.0 ng/g f.wt, respectively.

Human intake A daily intake of OP and NP from Italian seafood has been estimated to 0.05 and 12.2 μg/day and person, respectively [42]. Calculations were based on Italian consumption statistics with an adult seafood consumption of 31.8 g/day [46], and an average NP concentration of 384 ng/g f.wt and an average OP concentration of 1.57 ng/g f.wt, in 12 different species (the same amount of different species consumed) sampled along different areas of the Italian coast [41, 42]. A daily intake of NP from molluscs from the same area has also been roughly estimated [41]. For the general population, an assumed consumption of molluscs at about 20-40 g per person and day would correspond to a NP intake of 6-13 µg per person and day [41, 47]. High

14

consumers, consuming 150-270 g/molluscs a day would have an NP intake of 48-87 µg per person and day or approximately up to 1400 ng/kg bw. A worst-case intake of OP and NP from fish and mussels has been calculated in Germany [48]. Assuming a consumption of 300 g fish filet/ day with maximal concentrations of OP and NP found in fish from the German River Saar (2001), an intake of APs from fish was 3.1 µg/day and from mussels 0.8 µg/day (consumption of mussels not reported). In a German study, an estimated daily intake of NP from general food has been calculated [8]. Using German food consumption statistics, food samples from a typical German market basket were collected, representing food from all different groups of food. Based on concentrations of NP in food samples and daily consumption, an intake of NP from food was calculated to 7.5 μg/day and person. Using the same approach, an intake of NP for infants exclusively fed with breast milk or infant formula was estimated to 0.2 and 1.4 μg/day, respectively. In Swedish drinking water OP and NP have been measured to 0.02 μg/l and 0.29 μg/l, respectively [7]. Assuming a person drinks two litres of water per day containing measured Swedish water concentrations, a daily intake would be 0.571 ng/kg bw and 8.29 ng/kg bw for OP and NP, respectively. Time trends of OP and NP in breams (Abramis brama) of German rivers suggest that concentrations of APs decrease in fish [49]. Measured concentrations of OP in bream muscles from different locations along River Rhine ranged between 0.4 and 0.7 ng/g f.wt 1995, between 0.3 and 0.6 ng/g f.wt 2001. In River Elbe, concentrations of OP in bream ranged between 0.3 and 1.4 ng/g f.wt 1993, between 0.2 and 0.3 ng/g f.wt 2001. Concentrations of NP in bream were <11.0 ng/g f.wt (1995) and <4.2 ng/g f.wt (2001) in River Rhine, 2.8-13.3 ng/g f.wt (1993) and <3.0 ng/g f.wt (2001) in River Elbe. In breams from the more contaminated River Saar, maximum NP concentrations decreased from 112 ng/g f.wt 1994 to 9.8 ng/g f.wt 2001. In the same study, concentration of NP in common mussels (Mytilus edulis) from the North Sea was reported to decrease. The highest NP concentration in mussels was 9.7 ng/g f.wt in 1985, and below detection limit (2.0 ng/g) during 1998-2001 [48, 49]. OP and NP concentrations in mussels from the Baltic sea were 0.3 ng/g f.wt and below detection limit, respectively [48]. Since water plants bioconcentrate APs efficiently (BCF above 1000), plants grown on sewage sludge soil may be a possible source for human exposure. However, studies with barley have shown that the uptake in terrestrial plants is low [50]. In PVC-plastic, earlier used in food in food packing materials, NP concentration has been found up to 200 mg/kg [7, 51]. A possible migration of NP from PVC-plastic to food has been reported to 100 ng/g food [52]. NP is also present in food wrapping plastics as result of impurity and degradation of NPEs and trisnonylphenylphosphite (TNPP). NPEs are used as emulsifiers in food-packing material to prevent fogging, due to condensation of water [1]. The usual form of NPEs used is made by ethoxylating 1 mole NP with 4 mole ethylene oxide, where NP is present as impurity, of typically 0.1% NPE-4 (2.7% in PVC-plastic). NP is a residual impurity of acid hydrolysis of trisnonylphenylphosphite (used in food packing material as protection against degradation of UV-light) [1]. A total consumer exposure of NP from food due to NPE-4 and TNPP migration from food plastic materials has been estimated to 2 µg/kg/day [1].

15

NP and OP were measured in 25 adipose tissue samples from Swiss autopsies, from persons considered to be unexposed to APs [53]. Measured concentrations were 19-85 and 0.58-4.07 ng/g fat for NP and OP, respectively. Occupational exposure may occur during manufacturing and packing of APs. Workers have showed absorption of OP and BP based on urinary monitoring data [16, 54]. Exposure of BP may also occur from use or production of synthetic leather, adhesives and glues, germicides etc. [15, 55]. Consumers may be exposed to NP by the use of spermicidals, where a single dose would result in an exposure of more than 150 mg OP-9EO [7]. A worst-case scenario of individual human exposure to NP has been estimated [1]. A person living near a textile factory has an indirect exposure (drinking water and food) of 4.42 mg/kg bw/day. Exposure via hair dyes once a month, assuming that 0.02 ml of nonoxyl (present in hair dyes) may be metabolised to NP and a dermal exposure of 10 % with 10% adsorption, is estimated to 3 μg/kg per event or 0.1 μg/kg bw/day. Exposure to NP from food-contact materials is calculated to 2.0 μg/kg/day. Based on American residue and consumption data and a bioavailability of 10 %, an estimated intake from food plastic migration is 0.2 μg/kg bw/day. The use of pesticide products indoors with an air concentration of 60 μg/m3 for 20 minutes, gives an exposure of 21 μg via inhalation, 3.2 μg via dermal hand surface exposure, estimated exposure 0.35 μg/kg bw/day. Use of speciality paints (8 hours) gives an inhalation exposure of 1 ppm, a dermal exposure of 0.25 mg/cm2, or an estimated exposure of 2.0 mg/kg bw/day. A total exposure from above sources is 6.4 mg/kg bw/day.

Kinetics and metabolism of alkylphenols

Uptake Several studies of OP and NP on rats have showed that gastrointestinal tract absorption is the main exposure route. The bioavailability of OP and NP following oral administration seems to be low, about 10%. For humans occupational exposed to BP, uptake is mainly by inhalation [12]. Dermal uptake may occur if protective equipment is not used. In experiments with rats, t-OP (single oral dose, 50 or 200 mg/kg bw) had a bioavailability of 2- 12 % depending on the dose [56, 57]. Low plasma levels of OP have been explained by a first-pass effect, with extensive metabolism in the liver. Low skin penetration of NP has been observed in a human study [58]. An in vitro study with 14C-ring labelled NP applied to human skin (dermal dose 0.3 mg/cm2) showed an absorption and penetration of 0.1% and 4 % respectively, of administered dose. Inhalation data of APs are lacking. However, if inhalation occurs, bioavailability is high since no first pass effect takes place before APs can reach target organs [1].

Distribution and excretion APs are metabolised by glucuronation and sulphation in the liver [56]. As glucuronide conjugation of APs take place both on the aromatic hydroxyl group and the hydroxylated alkyl chain, this is considered the major metabolic pathway [59]. In mammalian liver, APs are hydroxylated by the P450 enzymes families CYP1A and CYP3A, to catechol structures [21]. Aromatic hydroxyl groups may be O-methoxylated by the catechol O-methyl transferase (COMT). Mono-glucuronide conjugates are formed by uridine diphosphoglucuronosyl-

16

transferase (UDPGT) enzymes [59]. Sulphated conjugates of hydroxylated phenols are formed by the enzyme family sulfotransferases with the co-factor 3´-phosphoadinosine-5´phosphatase (PAPS) [60]. Bilary excretion of glucuronide conjugates may lead to de-conjugation by ß-glucuronidases in the gut, leading to enterohepatic circulation. For t-OP, extensive enterohepatic circulation have been showed in female DA/Han rats [57]. 4-t-BP is rapidly metabolised and excreted as conjugates in urine. Radiolabelled BP was given intravenously to Wistar rats (single dose 1.2-10.4 mg/kg bw) and bile and urine were collected for four hours [61]. Total recovery was 91-93% of which 65-71% was excreted as glucuronide conjugate, 17-21 % as sulphate conjugate. In male Wistar rats given 14C- labelled BP (147 μg/kg bw/day) by gavage for 3 days, 72.9 and 26.7% of the administered dose were found in sampled urine and faeces, respectively [37]. In an in vitro study in Sprague Dawley hepatocytes, several metabolites of t-OP have been identified (Fig. 2). Four of five OP-metabolites (I, III, IV and V) were formed within 5 minutes by aromatic hydroxylation, forming catechols or conjugated catechol structures [59]. During these 5 minutes of incubation, 57% of the t-OP dose was metabolised. Within an hour, the metabolites were further oxidised in the alkyl chain to alcohols, or even further oxidised to glucuronide conjugates.

Fig. 2. Proposed metabolic pathways for t-OP in Sprague Dawley hepatocytes by Pedersen and Hill [59]. Bold arrows show major pathways, bracketed structures unidentified metabolites. GlcUA= glucuronide.

17

Glucuronation of the APs ethylphenol, BP, hexylphenol and NP have been studied in situ in Sprague Dawley rats [62]. Transport of alkylphenol glucuronide conjugates from liver to bile by MRP2 (multi resistant protein) was found to decrease with longer alkyl chain. For NP, most of conjugated NP remained in the liver after an hour of perfusion, suggesting a disturbed transport of NP-glucuronide across the canalicular membrane into bile. Other glucuronide conjugates of alkylphenols were rapidly transported to bile. Elimination half-life in blood of OP after a single intravenous injection (5 mg OP/kg bw) to Wistar rats was determined to approximately 5 hours [56]. In contrast, half-life in DA male rats was 31 hours in a similar study [57]. Repeated dosing of high amounts of OP did not result in increased tissue concentrations and low doses of OP seem to be removed efficiently due to a first pass effect [56]. In Wistar rats treated with OP in drinking water (8 mg/l, about 800 µg/kg) for 14 or 28 days, no OP was detected in the blood (detection limit 1-5 ng/ml) [56]. With the exception of some single animal in the 14 days exposure group, no concentrations of OP were found in fat, liver, kidney, brain, lung, testis or muscle. Oral administered NP in rats is eliminated in the faeces (80%) and in urine (20%) [63]. The high amount on elimination in the faeces may depend on metabolism and conjugation in the liver, but also on low uptake of NP in the gastro-intestinal tract. Elimination half-live of 4-NP in blood following gavage administration (50 mg/kg bw) has been calculated to 3.1 and 4.0 hours for male and female Sprague Dawley rats, respectively [64]. Transplacental transfer of 4-NP has been found in rats [64]. Sprague Dawley rats were given 50 mg/kg bw NP by gavage day 20 of pregnancy and fetuses were removed by caesarean section after 1-2 hours exposure. Concentrations of NP in fetal serum (3 litters) ranged from 4.9-8.0 µM, of which 13 % NP was present as active aglycone, the rest as conjugated NP. NP was also present in the brain of dams, where total NP concentration ranged between 12 and 20 µM. A human metabolism study of 4-NP has been made on two male volunteers, age 29 and 58 years [53]. One volunteer was given a single oral dose of 66 μg/kg bw NP, the other a single intravenous dose 14 μg/kg bw NP. After oral administration, blood levels peaked at about an hour with a concentration of 86µg/kg blood for conjugated NP (100 times higher than that of unconjugated NP). Following intravenous administration a maximum blood concentration was reached within 30 min, where conjugated and unconjugated NP was 0.6 and 0.2 μg/kg blood respectively. Comparing AUCs (Area Under Curve) for different administration routes, an oral bioavailability for unconjugated NP was suggested to about 20%. For both administrations, half-life in blood was 2-3 hours for the parent compound. Elimination routes of NP after oral administration were urine (10 %) and faeces (1.5 %) during a 56 hours collection period. Since only 11.5 % of NP was recovered, bioaccumulation in fat may occur. In rats, molecules weighing below 250 g/mol are to a larger extent excreted in urine, where molecules above 350 g/mol are to a lager extent excreted in faeces [60]. Metabolism studies of OP in rat (glucuronide and sulphate conjugate weighing 381 g/mole and 285 g/mole) show that faeces is the main elimination route [56]. In humans, conjugates with molecule weight up to 450-500 g/mole are mainly excreted in urine [65]. APs in humans are therefore expected to be eliminated mainly in urine, where enterohepatic circulation playing only a minor roll.

18

Toxic effects on laboratory animals and in vitro systems

Acute toxicity The acute oral LD50 of BP in rat, is about 4000 mg/kg [11, 12]. LD50 (oral) for Sprague Dawley rats have been reported to 5360 mg/kg for males 3620 mg/kg for females [66]. Observed clinical symptoms of toxicity in rats are sluggishness, unsteady gait, physical weakness, scruffy appearance and nasal discharge. LD50 (skin) in rabbit is about 2200 mg/kg and inhalation LC50 in rat above 5600 mg/kg [66]. Signs of inhalation toxicity in rats are mucosal irritation and respiratory distress. Two rats that died had dark red or purple discoloration of lungs and/or kidneys. Reported acute oral toxicity of OP in rats range between 1000 and 4040 mg/kg bw [67]. For mouse LD50 have been reported to 3210 mg/kg bw. Toxicity following intraperitoneal administration is much higher due to the lack of first pass metabolism, where LD50 in mouse have been reported to 38 mg/kg bw. Dermal toxicity in rabbits is above 2000 mg/kg bw. Signs of acute oral toxicity in rats are lethargy, diarrhoea, rough four, bloody nose and difficulties in breathing. For NP oral LD50 in rats are reported to range between 1200 and 2400 mg/kg for males and 1600-1900 mg/kg for females [1]. Clinical signs of toxicity are excessive salivation, diarrhoea and lethargy. At necropsy erosion of stomach muscle was seen. LD50 (skin) in rabbits is reported to 2031 mg/kg and oral LD50 in mice 307 mg/kg [68, 69]. Irritation Several reports have found BP moderate-highly irritating to skin of rabbits, and highly-corrosive irritating of the eye of rabbits [12]. Klonne and co-workers reported 4-t-BP to be irritating to the eye and skin of rabbits, with the eye irritation capable of causing severe corneal injury and severe conjunctive irritation [66]. In several reports, OP have been reported to be irritating to the skin of rabbits [67]. Others studies have reported OP to be not irritating, slightly irritating or corrosive to the skin of rabbits. OP has also been found irritating or highly irritating to the eye of rabbits. NP cause severe irritation of the eye of rabbits. Studies have reported conjunctive redness and lesion of the iris [1].

In vitro genotoxicity studies 4-BP was negative in the Salmonella test with or without metabolic activation in Escherichia coli, Saccharomyces cerevisiae, and in tests of chromosomal changes with cultivated cells from rat liver [70]. P-t-BP induced structural chromosome aberration in 6.5-12.0% of studied Chinese hamster lung cells (CHI/IU) with exogenous activation in all concentrations (ranging 0-0.05 mg/l) [71]. Continuous treatment (48 hours) at cytotoxic concentrations of 0.025 and 0.05 mg/l induced polyploidy with and without metabolic activation. OP9EO was negative in the following tests: Inhibition of unscheduled DNA synthesis in isolated mouse leucocytes [72], induction of DNA breaks in mouse lymphoma cells [73], and induction of chromosomal aberrations in Chinese hamster cell line [74]. Neither OP or NP caused any increase in mutations in HGPRT mutation assay in rat liver cell-line or any increase in unscheduled DNA repair in isolated rat hepatocytes [75]. NP potential

19

to induce mutations in HGPRT-locus was also tested in Chinese hamster V79 cells [76]. With an exposure period of 5 hours and concentration up to 2.5 μg/l (with metabolic activation) or 1.25 μg/l (without metabolic activation) NP was negative.

In vivo genotoxicity studies Considering in vitro studies of mutation and induction of chromosomal aberrations, OP and NP have a low genotoxic potential. Two in vivo tests studying NP effects on micronucleus in mice bone marrow were negative. NMRI strain mice (5 female and 5 male) received a single intraperitoneal dose of 50, 150 or 300 mg/kg [77]. Bone marrow was sampled after 24 hours for all dose groups, and also after 48 hours in the highest dose group. No effect on the ratio of polychromatic to normochromatic erythrocyte (PCE/NCE) was seen. In a similar study following oral administration (single dose 500 mg/kg), bone marrow sampling at 18, 48 and 72 hours, showed no effects on the ratio PCE/NCE for any of the sampling times [78].

Subchronic exposure Adult male Fischer rats (n= 6/group) were exposed to 4-t-OP in drinking water (0.02-0.035, 2.0-3.5 or 200-350 µg/kg bw /day) for 4 months [79]. For all treated groups, abnormal sperm development such as broken, coiled or bent sperm tails were common. In the highest exposure group, a decrease in number of epididymal sperm head per gram tissue and per organ was observed. No effect was seen on body weight, reproductive organ weights, luteinizing hormone (LH), follicle stimulating hormone (FSH), and testosterone serum concentration or testicular sperm number. A LOAEL based on sperm tail development was set to 20-35 ng/kg bw/day. Sprague Dawley rats (n=15/group/sex) were exposed to p-NP in diet containing 0, 200, 650 or 2000 ppm (approximately 0, 15, 50 and 140 mg/kg/day) for 90 days [80]. A reduction of mean total body weight was observed for both male and female rats at the highest dose level during the experiment. For the same dose level, males had increased mean kidney weight. After a recovery period of 4 weeks, kidney weight was decreased to the same level as controls. Females in the 2000 ppm group had reduced ovary weight. Other variables studied, including a number of organ weights, sperm analysis and estrus cycle length showed no significant change.

Chronic exposure and carcinogenic effects Male Fisher rats were treated with 150 mg/kg 4-t-BP by stomach tube for one week, thereafter with 1.07 g/kg bw BP in diet for 51 weeks [81]. The treatment with BP was given with or without an initiator, n-methyl-N´nitro-N- nitrosoguanidine (MNNG). For rats treated with BP and MNNG, an increase of carcinoma in situ (15/20) and forestomach carcinogenesis papilloma (8/20) and squamous carcinoma (15/19) were seen. Initiator alone had an effect of carcinoma in situ (13/19), papilloma (11/19) and squamous carcinoma (5/19). Results suggest that BP may have promoting effect on forestomach carcinogenesis. Newborn female Donryu rats were treated with p-t-OP (100 mg/kg/day) by subcutaneous administration every other day, from day of birth to postnatal day 5, or from day of birth to postnatal day 15 [82]. All females were also given a single injection of N-ethyl-N´-nitrosoguanidine (initiator) into the uterine horn at 11 weeks of age. At 15 months of age occurrence of total adenocarcinoma (different degrees of hyperplasia) in control rats was 6/23, in rats treated postnatal day1-5 18/28, and 8/22 in rats treated postnatal day 1-15. For treated rats during longer exposure period, malignity of uterine tumours was increased (measured as tumour differentiation and invasion in near and/or distant organs).

20

Male Fisher rats (n=12-17/group) were intraperitoneally given the initiator 3,2-dimethyl-4-aminobiphenol (DMAB) (25 mg/kg bw) every other week for 20 weeks and thereafter 4-n-OP (linear isomer) by diet in concentration of 10 or 100 ppm (highest dose about 4.4-5.6 mg/kg bw/day) for 40 weeks, starting one week after last dosing of the initiator [83]. No effect of administered OP was seen on the incidence of prostatic adenocarcinoma. Three weeks old male Fisher rats (n=20/group) were treated with p-NP in diet 0, 25, 250 or 2000 ppm (250 ppm about 35-45 mg/kg bw/day) for 3 weeks [84]. From week 7 of age, rats were treated subcutaneous with the initiator 3,2’-dimethyl-4-aminobiphenyl (DAMB) 50 mg/kg bw every other week for 20 weeks. An additional group (n=15) was treated with the highest NP dose without administration of initiator. No effects due to NP treatment were seen on incidence, multiplicity or area of neoplastic lesions in the prostate and seminal vesicles. Authors suggest that neonatal treatment of NP has no effects on rat prostrate carcinogenesis. Transgenic rats carrying the human c-Ha-ras proto-oncogene and non-transgenic rats were given a single dose of the initiator 7,12-dimethylben[a]anthracene (DMBA) (25 mg/kg bw) by gavage, and thereafter NP by diet (0, 10, 25, 100 or 250 ppm) for 8 weeks (transgenic females) or 20 weeks (non-transgenic females, transgenic and non-transgenic males) [85]. The number of animals was 7-10/sex/group. Female transgenic rats in the 10 ppm group had increased incidence of adenocarcinoma and total mammary tumour multiplicity and males in the 25 ppm group had increased sarcoma multiplicity, but not at higher dose levels. No dose-response effects were retrieved from this study.

Reproductive and developmental effects In vitro effects

Endocrine effects of APs have been showed in mammalian cells [5]. APs with the alkyl group in p (para) position are able to stimulate growth of both MCF-7 cells and human breast cancer cells [86]. Of the APs, OP is most potent, about 1000 times less potent than 17β-estradiol [5, 87]. The binding affinity of 4-t-OP relative 17ß-estradiol to bacterially expressed glutathione-S-transferase (GST)-estrogen fusion proteins from different species has been compared. To human fusion proteins, OP had about 10 000 times lower affinity than 17β-estradiol [87]. The difference in relative binding affinity ((IC50 17β-estradiol/ IC50 OP)·100, where IC50 is the concentration of OP required to displace 50% of 17β-estradiol from GST-estrogen fusion proteins) between mammals were small, 0.17 and 0.12 for mouse and man, respectively.

Short term studies Studies of BP are scarce. In one briefly reviewed study, BP was given to Sprague Dawley rats by gavage (0, 20, 60 or 200 mg/kg bw/day) for 44 days for males (14 days before mating to 14 days after mating), 14 days before mating to day 3 of lactation in females (number of animals not reported) [71]. No treatment related effect on pregnant females, lactating females or their offspring was seen. In an in vivo study of pigs exposed to 4-t-OP during pregnancy (subcutaneous 1 mg/kg bw for 63 days), OP did not exert estrogen-analogous induction of estrogen receptor α in the placenta (4-6 autopsies from one animal) [88]. No difference in placental morphology was seen compared to control.

21

Adult male Fisher rats (n=6/group) were treated with t-OP (1·10-5, 1·10-7 or 1·10-9 M) in drinking water for 6 days [79]. No effect on body and organ weight or serum concentration of LH, FSH and testosterone was seen. Female Wistar rats were exposed to 4-t-OP in drinking water 2 weeks before mating to gestation and/or from day 1 to day 22 after giving birth. Born male rats had reduced testis and prostate weight at a dose level as low as 12.9 µg/kg bw/day (100 µg/l in water) [89]. At the highest dose of 1000 µg/l in water, daily sperm production was reduced (other doses not evaluated). However, same authors failed to reproduce the result in a later study, where even the highest dosed failed to cause any significant decrease in testis weight or sperm production [90]. Female newborn Donryu rats were injected with 100 mg p-t-OP /kg bw subcutaneously every other day (totally 8 doses) from day of birth to postnatal day 15 [91]. Levels of FSH and LH in treated rats were lower during OP exposure than in controls rats (n=4-5/group). Also, normal hormone peaks in serum were absent. The treatment caused about 4 days earlier vaginal opening (n=21), persistent estrus (n=24) and uterine endometrial hyperplasia (n=4-5). None of the OP-treated rats showed regular estrus cycle during the study period. OP was given subcutaneous 20 or 80 mg three times per week (about 34 and 137 mg/kg bw/day) to 2 months old male Fischer rats for either 1 or 2 months (n= 6 or 5/group) [92]. Adverse effects on secretion of reproductive controlling hormones were seen. For rats treated with 80 mg OP, serum concentration of LH and FSH decreased after 1 or 2 months treatment, while prolactin concentration increased during both treatment periods. The same hormone effects were seen for the lower exposure group after 1 month of treatment, but not after 2 months. The higher OP dose also resulted in suppressed food consumption for both treatment periods, and increased spleen weight after 2 months treatment. A positive control of estradiol valerate (dose 8 μg) resulted in similar LH and FSH hormone levels as 80 mg OP, but about 2-4 times higher prolactin levels than rats treated with 80 mg/kg bw/day OP for 2 months. 19 Suffolk cross ewe lambs (n=3-6/group) were born to ewes given subcutaneous injections of OP twice weekly (1.0 mg/kg bw/day) during gestation and lactation [93]. Exposure to OP was either from day 70 of gestation to weaning (5 months of age), day 70 of gestation to birth, or birth to weaning. For all OP exposure groups, onset of puberty, measured as first progesterone rise followed by estrus, occurred about a month earlier than for control lambs. No effect was seen on FSH levels, interestrus intervals or ovarian follicle size and number. Neonatal exposure to APs may affect reproductive development. Newborn female Sprague Dawley rats (n=6-8/group) were injected intraperitoneally with OP or NP (5 or 50 mg/kg bw/day), day 1 to day 10 after birth [94]. For the highest exposure of OP, the day of vaginal opening occurred several days earlier compared to controls. Also, weight increases of ovaries and uterus were observed. Reproductive failure, measured as percent rats that failed to ovulate, was observed to 36 % and 85 % for the dose groups 5 and 50 mg/kg bw/day OP. For the higher dose of NP reproductive failure was 18 %. Also, a persistent estrus cycle after vaginal opening and absence of LH peaks, were seen in all animals treated with 50 mg/kg bw/day OP. No effects on onset of puberty or reproductive tract development were seen for NP. NP (100 mg/kg/day orally for 10 days) increased the mammary gland differentiation and cell proliferation in Alpk rats (12 exposed animals) [95]. In female Noble rats (n= 10/group) p-NP

22

(subcutaneous implanted minipump, 0.075 or 53.2 mg/kg/day for 11 days) had no effect on mammary gland differentiation and cell proliferation [96]. In contrast, an increased cell proliferation on mammary epithelial cells as well as altered cell-cycle length were observed in Noble rats with the same study design at an even lower dose, 0.073 mg/kg/day [97]. Adult male Sprague Dawley rats (n=20/group) were treated with p-NP intragastrically (100, 250, 400 mg/kg bw/day) for ten weeks, a period with at least one spermatogenetic cycle [98]. Survival in the two higher dose groups was only 10-25%. In the lowest exposure group, seminiferous tubule diameter was decreased. Body weight, testicular weight, epididymal weight and sperm count were unaffected at this dose level. 10 female Sprague Dawley rats were exposed to p-NP (100, 250 or 400 mg/kg bw/day) by gavage from day seven of gestation to weaning of their litters. Born male pups were further exposed (n=20/group) at the same dose level as their mothers until 10 weeks of age [99]. No pups were born from females exposed to the highest dose. Pups of the lowest dose group showed decreases in body and testis weight and in seminiferous tubule diameter. In the two higher doses groups, decreased epididymis weight and decreased epididymal sperm count were also observed. No physical or behaviour abnormities was seen in dams or pups, neither malformations nor stillbirths. Newborn male Sprague Dawley pups (n=3-4/group) were given NP (8 mg/kg bw/day) intraperitoneally for 18 days, starting day 1, 6 or 13 after birth [100]. Exposure during day 1-18 resulted in decreases in testis, epididymis, seminal vesicle and ventral prostate weight on body weight basis. Similar effects were seen in pups exposed day 6-24, but not in pups exposed day 13-30. A critical development period was identified to the two first weeks of life. Treated newborn pups developed abnormal decent of testis resulting in cryptochidism (failure of testis to descend from abdomen). NP-treated pups were also shown to have reduced reproductive capacity. More than half of treated males were infertile or less fertile compared to control animals, when mated with unexposed females. When intraperitoneally administering NP 0.08, 0.8 or 8.0 mg/kg bw/day daily for 15 days (postnatal day 1-15), a dose dependent decrease of testis, epididymis and seminal vesicle weight and an increase in prostate weight was seen for the two highest doses (n=3/group).

Multigenerational studies No study on BP has been found. In a two-generation study, adult Sprague Dawley rats (n=30/sex/group) were exposed to p-t-OP in the diet (0, 0.2, 20, 200 or 2000 ppm) for 21 days before mating, during mating (10 days), gestation (21 days), and lactation (21 days) [101]. Same exposure design was used for F1 and F2 generation, where F2 rats were sacrificed day 111 after birth. No effects on sperm measurements, estrus cyclicity or reproductive organs were seen in adult rats in any generation. At the highest dose level, decreased body weight or decreased body weight gain was seen for both sexes in all generations. For the highest dose, F0 rats had decreased uterus weight (absolute and relative), and pups in F1 litters had decreased body weight during the latter portion of lactation (probably due to diet exposure). NOAEL for systematic and postnatal toxicity was 200 ppm (about 15 mg/kg bw/day) and for reproductive toxicity above 2000 ppm (about 149 mg/kg/day). In a three-generation study, t-OP was given daily to Landrace/ Yorkshire breeds sows (0, 10 or 1000 µg/kg bw/day) by subcutaneous injections from day 23 after insemination to day 85

23

of pregnancy [88]. In the parental generation (n=3/group), length of gestation increased in a dose dependent manner, where increased period of gestation is a possible sign of delayed fetal development. F0 sows treated with 1000 µg/kg OP (lower dose group not examined) had induced cervical cell proliferation. First estrus cycle occurred earlier in the 10 µg/kg group in the F1 generation compared F1 controls, but not in the highest dose group. Reduced litter size was also observed for born F2 piglets in a dose dependent manner. No effect on sex ratio, number of stillborn piglets or birth weight were observed in any generation, neither any effects on morphology of sex organs in F1 generation. In a three-generational study on Sprague Dawley rats (n=30/sex/group), the lowest NP dose causing adverse effects on kidneys function and sperm morphology was 200 ppm in diet. For all doses levels (200, 650 and 2000 ppm), increased kidney weight and renal tubular dilatation was found in all generations [102]. For the two highest dose levels, day of vaginal opening occurred earlier than in control rats (in F2 approximately 2 and 6 days). In F1 generation, uterus weight increased with 14% and 42% for the two highest dose levels, respectively. Longer estrus cycle was found in F1 and F2 generations exposed to 2000 ppm. No changes in testicular or testicular epididymis weight where seen. An increase in abnormal sperm morphology was found for all doses in F2 generation, and for the two highest doses also a decreased epididymal sperm density. Calculated dose corresponding to 200 ppm (mean intake during non-reproductive phases) gives a LOAEL of 15 mg/kg bw/day [103]. Pregnant female Sprague Dawley rats (n=15/sex/group) exposed to p-NP in soy-free diet (0, 5, 25, 200, 500, 1000 or 2000 ppm) from gestation day 6 [104]. Born pups were weaned day 21 after birth, and thereafter exposed to NP at the same dose levels as their mothers. No effect on gestation time, litter size, sex ratio or pup birth weight was seen for any of the doses. A decreased food consumption and decreased body weight gain during pregnancy were seen for the two highest exposure groups. For all F1 rats exposed to 2000 ppm NP, severe polycystic kidney disease was observed as numerous large cystic tubules, often uniformly distributed from one pole to another. In the 1000 ppm group, 67% of the males and 53% of the females had mild to moderate polycystic kidney disease. No adverse effects were seen in rats exposed to 500 ppm, or at lower dose levels. In a two-generation study on Sprague Dawley rats (n=10/sex/group), pregnant females were exposed to NP in diet 0, 25, 500 or 2000 ppm for 65 days, starting day 7 of gestation [105]. Pups were weaned postnatal day 22, and exposed to NP in diet until sacrifice postnatal day 65. In the 2000 ppm group, F0 females had decreased body weight and increased percentages of CD4+ T-cells and NK-cells. Males in the F1 generation at the highest dose level had increased spleen weight/body weight and increased spleen cell number. A decrease in relative number of NK-cells to number of spleen cells was observed at all dose levels for F1 males, but not in a dose dependent manner. Females in the F1 generation in the 2000 ppm group had increased relative spleen and thymus weights and decreased body weight. For F1 females, a dose dependent increase of NK-cell activity was also seen (two highest dose groups significant different from controls), when expressed as lytic units (number of splenocytes required to kill 10% of target cells) per spleen. Authors suggest that NP exposure in diet may alter immunological parameters in rats, mainly in the F1 generation, and that immunological changes possibly is gender-specific. A two-generation study, CD-1 mice (n=6/sex/group) were treated with p-NP (50 or 500 μg/l) in drinking water [106]. F0 generation were exposed for 4 weeks, F1 generation, during gestation and lactation up to adulthood, where males and females were sacrificed 100 days

24

and 130 days old, respectively. For males in F1 generation, increased absolute kidney weight and decreased relative liver weight were observed in dose dependent manner. Also, spermatogenetic changes were seen as decreased seminiferous tubule diameter and epithelium thickness for both doses levels. Acrosomal damage to mouse sperm was seen in both generations. For females, decreased relative kidney weight was seen for both doses levels in the parent generation. A decrease in second litter size in F2 generation was observed. No effect on sex ratio was seen.

Environmental toxicity In aquatic systems, the acute toxicity of APs is high, and toxicity increases with the length of the alkyl chain, see Table 3. Both OP and NP are highly toxic to fish, as well as highly toxic/ toxic to invertebrates [16, 38, 107]. Table 3. Acute toxicity of BP, OP and NP to aquatic organisms. Species Substance Endpoint/ Results (mg/l) Reference Test conditions Algae Scenedesmus subspicatus 4-t-OP 72h EC50 growth rate 1.1 [67] S. subspicatus NP 72h EC10 biomass change 0.0033 [108] S. subspicatus NP 72h EC50 growth rate 0.323 [108] Selenastrum capricornutum 4-t-BP 72h EC50 biomass change 22.7 [71]

Chlorella vulgaris 4-t-BP 6h EC50 biomass change 22-34 [109]

Chlorella pyrenoidosa NP 24h LC50 growth reduction 0.025-7.5 [110]

Invertebrates Water flea, (Daphnia magna) 4-t-BP 48h EC50 immobilization 3.4 [111] D. magna 4-t-OP 48h LC50

1. 0.17 [67]

D. magna NP 48h EC50 1. 0.085 [107]

Painted shrimp, (Hyallella azteca) NP 96h EC50 loss of mobility 0.0207 [107] Shrimp, (Crangon septemspinosa) 4-t-BP 96h Lethality threshold 1.9 [111] C. Septemspinosa 4-t-OP 96h LC50

1. 1.1 [36] C. Septemspinosa NP 96h LC50

1. 0.3 [36] Mysid shrimp, (Mysidopsis bahia) NP 96h LC50/ EC50 0.043 [4] Soft-shelled clams, (Mya arenaria) 4-t-BP 96h Lethality threshold > 9 [111] Fish Fatheaded minnow, (Pimephales promelas) 4-t-BP 96h LC50 5.1 [112]

P. promelas 4-t-OP 96h LC501. 0.25 [67]

P. promelas NP 96h LC501. 0.13 [38]

Salmon, (Salmo salar) p-NP 96h LC501. 0.13 [36]

Rainbow trout, (Oncorhyncus mykiss) 4-t-OP 96 h LC50

1. 0.12 [16]

O. mykiss NP 96 h LC501. 0.22 [107]

Age 45 days post hatch Killifish, (Fundulus heteroclitus) 4-t-OP 96h LC50 embryos 18.7 ·10-6M [113] F.heteroclitus NP 96h LC50 embryos 5.0 (24.7 ·10-6M) [113] 1. Flow trough conditions. 2. Static conditions.

25

Several studies have reported endocrine effects of APs in aquatic organisms. In rainbow hepatocytes, APs mimic 17β-estradiol by binding to the estrogen receptor [5]. In a competitive manner, OP and NP were able to displace17β-estradiol from its receptor. APs can cause feminising effects in male fish, such as increased plasma vitellogenin concentration and abnormal gonadal development [6, 114]. OP have been reported to induce vitellogenin production in Japanese medaka (Oryzias latipes) [6], rainbow trout (Oncorhynchus mykiss) [114], flounder (Platichthys flesus) [115] and in eelpout (Zoarces viviparous) [116]. A LOEC (vitellogenin production) for OP is 4.8 μg/l (rainbow trout) [114] and 10 μg/l NP/l (fatheaded minnow, Pimephales promelas) [117]. Vitellogenin production in Japanese medaka has been correlated to adverse reproductive effects [6]. Male vitellogenin production after OP exposure (measured concentration 20, 41, 74 or 230 μg/l) during 21 days, was correlated with decrease in number of fish eggs, % fertilized eggs and embryo survival when males were mated with unexposed females. The number of produced eggs decreased to almost half in all exposure groups. OP and NP exposure have developmental effects in killifish (Fundulus heteroclitus) embryo and larvae. Embryos exposed to OP or NP in a water concentration of 25 μg/l had abnormalities in major structures, resulting in mortality before hatch. In the water concentration of 10 μg/l (day 10 post fertilization), abnormal larvae development such as lack of complete tail fins, spinal curvatures, lethargic swimming behaviour and/or inability to feed was seen for both OP and NP.

Human toxicity studies Single persons have been observed to develop allergy contact dermatitis to different APs [9]. Para-t-BP catechol and t-BP are known to cause depigmentation of the skin [118]. Several studies have reported workers with depigmentation of skin (as a disorder named vitiligo) due to handling of a mixture of BP, formaldehyde and derivates. In a Russian factory producing 4-t-BP and p-t-BP-formaldehyde resins, 23 of 52 workers had skin depigmentation after a year or longer exposure [12, 119]. In Germany 23 workers (total number of workers not reported) handling BP were observed to have skin depigmentation on hands and arms after 2 months up to 2 years exposure [120]. In an Austrian factory, ten male workers exposed to a mixture of BP, formaldehyde and derivates developed vitiligo after 10 months to 7 years of exposure [121, 122]. BP may also induce enzymatic activity in the liver [121, 122]. Enlarged livers were observed in 4 of 10 workers with developed vitiligo due to handling a mixture of BP and formaldehydes and derivates (2 months to 2 years exposure). In two of the workers with enlarged liver, an increased liver activity measured as sulfobromophtalein (BSP) clearance was seen. In one of the vitiligo patients, an enlarged spleen was observed. In vitro germ cells from human fetal gonads (6-9 weeks old) were cultured for 3 weeks, exposed to 1 μM 4-OP in medium [123]. Mitotic index (number of cells in mitosis/100 cells) and number of pre-spermatogonia were reduced in testicular tissues. No effect was seen on ovarian tissue.

Risk assessments and regulations A risk assessment of 4-t-OP has been made by the Toxicological Programme, Karolinska institutet [13]. In a drinking water study on rats, a LOAEL based on sperm tail abnormalities

26

was determined to 20 ng/kg bw/day [79]. A default safety factor of 100 was adjusted to 40, motivated by interspecies variation between human and rat binding affinity of OP to ER (×4) and intraspecies variation in sensitivity (×10). A TDI was suggested to 0.5 ng/kg bw/day. A German risk evaluation of NP estrogenic potential, concluded no risk for non-occupationally humans [124]: A daily intake of NP from 1 kg food contaminated by migration from PVC-plastic (≤0.1 mg/kg f.wt) [52], 100 g duck or fish (0.02-0.3 mg/kg f.wt) [40], and 3000 g drinking water (≤0.002 mg/l) [125, 126], was estimated to less than 0.16 mg per person. Comparing the difference in estrogenic potency of NP with 17β-estradiol (NP about 1000 times less potent), and a calculated blood concentrations in humans (using data from a human kinetic study [53]), a margin of safety (MOS) was estimated to 3000. Comparison of the calculated NP intake (0.16 mg/ person/day) to a rat NOAEL of 50 mg/kg [80], or a calculated organ concentration of 25 pg/g NP (assuming rapid fat distribution and ignoring elimination) with effects of 17-β-estradiol at 100 ng/g cell gave an even higher margin of safety. Workers manufacturing or using NP or NP-intermediates may be at high risk for repeated dose toxicity and reproductive system effects of NP [1]. Following inhalation exposure during 8 hours at a NP concentration of 0.1 ppm, assuming 10 m3 inhaled air and 100 % absorption, a daily exposure for workers was estimated to 0.13 mg/kg bw. In rats, LOAEL for renal toxicity was found to 15 mg/kg bw/day [103], and a NOAEL for reproductive effects (estrus cycle length, day of vaginal opening, ovarian weight and sperm/spermatid count) was 15 mg/kg bw/day [103]. Assuming oral bioavailability of 10 %, LOAEL and NOAEL were reduced by a factor 10 for comparison of inhalation or dermal exposure effects (100 % absorption). Comparing reduced LOAEL and NOAEL (1.5 mg/kg bw/day) with estimated exposure (0.13 mg/kg bw/day) the MOS was about 10 [1]. A TDI of NP has been reported by the Danish Institute of Safety and Toxicology [127]. A LOAEL for increased incidence of renal tubular degeneration and/or dilation in a three-generation study on rats was 15 mg/kg bw/day [103]. A total safety factor was set to 3000 motivated by: Humans are more susceptible than animals (×10), the most sensitive individuals should be protected (·10), LOAEL is used and carcinogenic and genotoxic data are lacking (×30). From above data, TDI was calculated to 5000 ng/kg bw/day. Authors also suggest that an existing limit value for phenols in drinking water 0.5 µg/l, offers adequate protection against adverse health effects of NP. Present study

Aim of study Literature data show that OP and NP are present in food, especially in fish and vegetables. The aim of the present study is to quantify OP and NP levels in fish from Swedish waters. Based on these results an evaluation of the potential human health risk after fish consumption of OP and NP is performed.

Materials and methods

Fish samples Data on Swedish fish concentrations of APs were derived from a report by Hajslova [128]. Samples of perch (Perca fluviatilis) and artic char (Salvelinus salvelinus) from Southern and Northern Vättern were collected by Vättenvårdsförbundet, bream (Abramis brama) from

27

Viskan River by Magnus Engwall (Örebro university), and rainbow trout (Oncorhynchus mykiss, farmed in Swedish waters) was bought in grocery stores by NFA (National Food Administration) [10]. Perch samples from Lake Bysjön, Lake Hjärtsjön and Kvädöfjärden were obtained from the Swedish Museum of Natural History. Fish samples were analysed by the Institute of Chemical Technology in Prague, using gas chromatography/mass spectrometry with selective ion motoring (GC/MS-SIM) [129]. Data on herring Clupea harengus)( , flounder (Platichthys flesus) and one perch sample (Lake Ringsjön) were taken from a Swedish screening study by IVL [20]. The number of fish samples analysed for OP and NP are presented in Table 4. Pooled perch samples analysed from Lake Bysjön, Lake Hjärtsjön, and Kvädöfjärden reported by Hajslova (totally 30 samples) consisted of 2-3 individual fishes per sample, bream samples from River Viskan consisted of 5 individual fishes per sample and the rainbow trout sample consisted of 15 fish samples [128, 129]. Pooled herring samples consisted of 5 individual fishes and samples of flounder and perch (Lake Ringsjön) consisted of 7 individual fishes [17, 20]. Intake calculations were based on median of measured concentrations of APs in fish, expressed as ng/g f.wt. For concentrations of AP in fish below detection limit, half of detection limit was used. Detection limit reported by Hajslova were 0.12 ng/g f.wt for both OP and NP [128], and detection limit for OP in flounder and herring reported by IVL was 0.30 ng/g f.wt [20]. Samples of bream (Viskan River) and one perch sample (Lake Ringsjön) were from polluted areas influenced by wastewater treatment and were therefore only included in worst-case intake calculations [129, 130]. Table 4. Concentrations (ng/g f.wt) of octylphenol (OP) and nonylphenol (NP) in fish from Sweden. Species Number1. Mean SD Median Max 95 percentile OP Perch (Perca fluviatilis) 40 0.267 0.320 0.15 1.67 0.7805 Artic char (Salvelinus salvelinus) 10 1.40 0.825 1.205 3.11 2.696

2.Bream (Abramis brama) 3 7.42 6.40 10.5 11.7 - Herring (Clupea harengus) 3 <0.30 - <0.30 <0.30 - Flounder (Platichthys flesus) 1 <0.30 - <0.30 <0.30 - Rainbow trout (Oncorhynchus mykiss) 1 <0.12 - <0.12 <0.12 - NP Perch (P. fluviatilis) 30 1.34 1.35 0.985 5.11 4.21 Perch (P. fluviatilis)2. 1 15 - 15 15 - 1. Refers to the number of samples. For the number of individual fishes in each pooled sample, see text. 2. Samples from polluted areas, bream from Viskan River, perch from Lake Ringsjön (IVL-data).

Calculations and statistics Calculations of intake 4-t-OP and 4-NP from fish were based on median concentrations in Swedish fish, and on Swedish fish consumption statistics from a report by NFA, named Riksmaten 1997/98 [131]. Intake of APs from fish was calculated for three exposure groups, women aged 17-40 (277 individuals), women 41-75 (350 individuals) and men 18-74 (573 individuals). The mean consumption of fish in Sweden was 29.9, 26.2, and 31.6 g/day for women 17-40, women 41-75 and men 18-75, respectively [131]. Four scenarios of OP exposure and two scenarios of NP exposure from fish intake were calculated (Table 5). Scenario 1 is a daily intake of OP from freshwater fish only. Scenario 2 is based upon measured concentrations of OP in freshwater fish, salmonids, herring and

28

29

flatfish (bream and perch samples from polluted areas excluded). Scenario 3 estimates intake of OP from all fish, including fish species from scenario 2 as well as cod, canned herring, other canned fish, fish dishes and eel. As an approximation of unknown OP concentrations in fish when calculating OP intake after all kind of fish consumption, concentration of eel was set to same value as for rainbow trout (0.06 ng/g f.wt), salmon was given the median concentration of artic char, and all other fish species the same concentration as low fat fish (flounder 0.15 ng/g f.wt). Scenario 4 is a “worst-case” intake of OP including all fish species in scenario 3, with all fish species having the highest measured concentration (bream in polluted area, 10.5 ng/g f.wt). For NP, only data on perch was available. A daily intake of NP from freshwater fish only was calculated (scenario 5), where perch from a polluted area was excluded. Also a “worst-case” intake of NP was calculated (scenario 6), where all concentrations were set to the highest measured concentration (pooled perch samples from polluted area, 15 ng/g f.wt). For all scenarios, mean, standard deviation, median, min, max, 5 percentile and 95 percentile are presented as ng/person and day, and on body weight basis in Table 6 and 7. In Fig. 3, daily intake in different scenarios is presented as mean and standard deviation on body weight basis. Frequency distributions of OP and NP intake are presented in Fig. 4 for women (17-41 years of age) on body weight basis. Statistics for comparisons of AP concentration in fish from different parts of Vättern, and between exposure groups were performed with Mann-Whitney test.

30

Table 5. Overview of fish species included in calculations of different scenarios. All fish concentrations are median values expressed as ng/g f.wt. Scenario Cod Flatfish Canned Other Fish Baltic Herring Salmon Salmon Baltic Freshwater Eel (flounder) herring canned dishes herring/ (farmed) (self salmon fish (perch) fish (58% cod) smoked caught) herring 11. 0.15 21. 0.15 0.15 0.15 0.06 0.06 0.06 0.15 31. 0.15 0.15 0.15 0.15 0.087 0.15 0.15 1.205 1.205 1.205 0.15 0.06 41. 10.5 10.5 10.5 10.5 10.5 10.5 10.5 10.5 10.5 10.5 10.5 10.5 52. 0.985 62. 15 15 15 15 15 15 15 15 15 15 15 15 1. Scenarios with calculated OP intake from fish. 2. Scenarios with calculated NP intake from fish.

31

Results OP and NP were detected in most samples of freshwater fish, but not at all in marine fish species. Generally, concentrations in Swedish freshwater fish of NP were higher than those of OP, with median concentrations of NP in perch about 6 times higher than those of OP (see Table 4). In the southern part of Vättern, perch had higher OP concentrations than in the northern part (p<0.05), indicating a higher pollution in Southern Vättern. No difference in OP concentrations in artic char between Southern and Northern Vättern was found (p>0.05). Fat freshwater fish from Lake Vättern such as artic char had higher OP concentration than perch from the same lake (p<0.05). Fish samples from polluted areas had much higher concentrations of OP and NP than fish samples from other areas. The NP concentration in perch from Lake Ringsjön (15 ng/g f.wt) had about 100 times higher median concentration than perches sampled in Lake Bysjön, Lake Hjärtsjön and Kvädöfjärden. Mean intake from freshwater fish in scenario 1 ranged between 0.0863 and 0.148 ng/day for adults, or between 0.00131 and 0.00184 ng/kg bw/day (Table 6. and 7.) An average intake of OP from fish species with known concentrations in scenario 2 was 0.80 –1.23 ng/day. In the same scenario, mean intake of OP ranged between 0.011 and 0.0186 ng/kg bw/day. Using scenario 3, an average estimated intake of OP from all fish ranged between 6.11 and 7.51 ng/day. On body weight basis, the mean OP intake was 0.093-0.115 ng/kg bw/day, where 95 % of adults had an exposure below 0.360 ng/kg bw/day. For scenario 4 and 6, mean of calculated worst-case intake from fish ranged between 206 and 235 ng/day for OP, and between 294 and 336 ng/day for NP. On body weight basis, women of age 17-40 had the highest average AP intake with 3.66 and 5.23 ng/kg bw/day for OP and NP, respectively. In the same scenarios, women (aged 41-75) had the highest individual exposure with maximum OP and NP intake of 45.4 and 64.9 ng/kg bw/day, respectively.

32

Table 6. Calculated intake of OP and NP from fish, expressed as ng/day and person for women of age 17-40 years (n=277), women of age 41-75 years (n=350) and men of age 18-75 years (n=573). Scenario 1-4 estimates the intake of OP, scenario 5 and 6 the intake of NP. For description of scenarios, see Table 5. Women 17-40 Women 41-75 Men 18-75 Scenario 1 Mean ±SD 0.0863±0.189 0.131±0.320 0.148±0.406 Median 0 0 0 Min 0 0 0 Max 1.25 2.50 6.69 5 Percentile 0 0 0 95 Percentile 0.156 1.25 1.25 Scenario 2 Mean ±SD 0.809±0.822 1.23±1.44 0.889±0.970 Median 0.488 0.753 0.618 Min 0 0 0 Max 5.00 17.5 9.17 5 Percentile 0 1.25 0.045 95 Percentile 2.08 3.50 2.73 Scenario 3 Mean ±SD 6.11±5.69 7.51±8.32 7.41±7.75 Median 4.84 4.80 5.94 Min 0 0 0 Max 55.3 68.3 109 5 Percentile 0.567 1.54 0.702 95 Percentile 16.2 22.5 17.6 Scenario 4 Mean ±SD 235±161 232±203 206±165 Median 211 191 177 Min 0 0 0 Max 1230 2270 1290 5 Percentile 35.2 40.4 20.8 95 Percentile 545 557 461 Scenario 5 Mean ±SD 0.567±1.24 0.862±2.10 0.970±2.67 Median 0 0 0 Min 0 0 0 Max 8.21 16.4 43.9 5 Percentile 0 0 0 95 Percentile 1.03 8.21 8.21 Scenario 6 Mean ±SD 336±230 333±290 294±236 Median 300 290 252 Min 0 0 0 Max 1755 3246 1850 5 Percentile 50.3 57.8 29.7 95 Percentile 779 776 658

33

Table 7. Calculated intake of OP and NP expressed as ng/kg bw/day for women of age 17-40 years (n=277), women of age 41-75 years (n=350) and men of age 18-75 years (n=573). Scenario 1-4 estimates the intake of OP, scenario 5 and 6 the intake of NP. Significant difference from other exposure groups is shown with different superscripts (p<0.05). For description of scenarios, see Table 5. Women 17-40 Women 41-75 Men 18-75 Scenario 1 Mean ±SD 0.00131±0.0027a 0.00196±0.00486b 0.00184±0.00498 Median 0 0 0 Min 0 0 0 Max 0.0208 0.0439 0.0787 5 Percentile 0 0 0 95 Percentile 0.0030 0.0143 0.0132 Scenario 2 Mean ±SD 0.0125±0.0128a 0.0186±0.0245b 0.0110±0.0116a

Median 0.00781 0.0111 0.00778 Min 0 0 0 Max 0.0833 0.350 0.0905 5 Percentile 0 0.00187 0.000572 95 Percentile 0.0343 0.0507 0.0343 Scenario 3 Mean± SD 0.0958±0.0922 0.115±0.130a 0.0939±0.101b Median 0.0734 0.0744 0.0713 Min 0 0 0 Max 0.922 1.09 1.36 5 Percentile 0.00792 0.0228 0.00767 95 Percentile 0.260 0.360 0.233 Scenario 4 Mean± SD 3.66±2.50a 3.55±3.45a 2.59±2.11b

Median 3.36 2.78 2.24 Min 0 0 0 Max 16.6 45.4 19.0 5 Percentile 0.486 0.597 0.262 95 Percentile 8.32 8.90 5.99 Scenario 5 Mean± SD 0.00858±0.0178 0.0129±0.0319 0.0121±0.0327 Median 0 0 0 Min 0 0 0 Max 0.137 0.288 0.517 5 Percentile 0 0 0 95 Percentile 0.0197 0.0942 0.0866 Scenario 6 Mean± SD 5.233.57a 5.074±4.93a 3.71±3.02b

Median 4.80 3.97 3.19 Min 0 0 0 Max 23.7 64.9 27.2 5 Percentile 0.694 0.853 0.374 95 Percentile 11.9 12.7 8.56

Women of age 17-40 in scenario 4 and 6 had the highest average and median intake of OP and NP from fish (Table 7), but there was no significant difference in OP or NP exposure compared to women of age 41-75 (p>0.05), see Fig 3. Although, the difference in exposure between women and men in scenario 4 and 6 was highly significant (p< 0001), little difference in mean exposure levels (1-1.5 ng/kg bw/day) were seen between groups.

Worst-case intake of OP from fish

women 17-40 women 41-75 women men 0123456789

1011

TDI= 0.5TDI (men)=0.067

TDI (women)=33.3

*

ng/k

g bw

day

and

pers

on

Worst-case intake of NP from fish

women 17-40 women 41-75 women men 01

23

45

67

89

10

11

T DI= 5000T DI=50 000

*

ng/k

g bw

/day

and

per

son

A. Fig. 3. Worst-case intake ofOP and NP (mean ± SD)from fish for women of age17-40 years (n=277), womenof age 41-75 years (n=350)all women (n=627) and men(n=573). *Significantdifferent from the intakeamong women, p<0.05.An earlier suggested TDI forOP of 0.5 ng/kg bw/day, andTDI suggested in presentstudy of 0.067 ng/kg bw/day(men) and 33 ng/kg bw/day(women) are shown in Fig3A. Earlier suggested TDI forNP of 5000 ng/kg bw/dayand TDI suggested in presentstudy 50 000 ng/kg bw/dayare expressed in Fig.3B.

B.

Frequency distributions for different intake scenarios of OP and NP from fish for women of age 17-41 are presented in Fig. 4. Generally, frequency distributions were skewed towards the lower end of the distribution. A few outliers had a high AP exposure due to high fish consumption, above 70 g/day (one individual about 170 g/day). The most important exposure source in scenario 3 was by intake of farmed salmon, using data from artic char, Lake Vättern. In worst-case scenarios 4 and 6, cod, fish dishes, farmed salmon and flatfish were important sources for high individual exposure. In scenario 2, a somewhat higher variation in daily intake of OP was due to individual variation in fish species consumed. When all fish species was included in scenario 3, variation in OP intake decreased

34

Scenario 4

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

5.00

5.50

6.00

6.50

7.00

7.50

8.00

8.50

9.00

9.50

10.0

010

.50

11.0

011

.50

12.0

012

.50

13.0

013

.50

14.0

014

.50

15.0

015

.50

16.0

016

.50

0.02.55.07.5

10.012.515.017.520.022.525.027.530.032.5

TDI= 0.5

T DI (women)= 33.3

Intake of OP ng/kg bw/day

freq

uenc

y (n

r)

Scenario 60.

001.

002.

003.

004.

005.

006.

007.

008.

009.

0010

.00

11.0

012

.00

13.0

014

.00

15.0

016

.00

17.0

018

.00

19.0

020

.00

21.0

022

.00

23.0

024

.00

0

10

20

30

40

T DI= 5000T DI= 50 000

Intake of NP (ng/g bw/day)

freq

uen

cy (n

r)

Fig. 4. Frequency distributions of intake of OP and NP (ng/kg bw/day) from fish for women of age17-41. An earlier suggested TDI for OP of 0.5 ng/kg bw day, and a TDI for women suggested inpresent study of 33.3 ng/kg bw/day shown in Fig. 4B and C. A. OP intake from fish speices withknown concentration (scenario 2). B. OP intake of OP from all fish (scenario 3). C. Worst-case intakeof OP from all fish (scenario 4). D. Worst-case intake of NP from all fish (scenario 6). An earliersuggested TDI for NP (5000 ng/kg bw/day) and TDI for NP suggested in present study (50 000 ng/kgbw/day) are shown in graph.

Scenario 20.

000

0.00

50.

010

0.01

50.

020

0.02

50.

030

0.03

50.

040

0.04

50.

050

0.05

50.

060

0.06

50.

070

0.07

50.

080

0.08

5

0

10

20

30

40

50

60

70

80

90

Intake of OP (ng/kg bw/day)

freq

uenc

y (n

r)

Scenario 3

0.00

00.

050

0.10

00.

150

0.20

00.

250

0.30

00.

350

0.40

00.

450

0.50

00.

550

0.60

00.

650

0.70

00.

750

0.80

00.

850

0.90

0

0102030405060708090

100110120

TDI= 0.5

TDI(women) =33.3

Intake of OP (ng/kg bw/day)

freq

uenc

y (n

r)

C.

A. B.

D.

35

36

Derivation of TDI Because of few available data, no conclusion can be drawn on BP. BP has been found in seafood in Singapore in concentrations of 6.5-24.0 ng/g f.wt, about the same levels as OP [28]. One short term toxicity study of BP on rats showed no adverse effects at dose levels up to 200 mg/kg bw/day [71]. In an in vitro genotoxicity study according to OECD TG-guidelines, p-t-BP induced chromosome aberrations and polyploidy in Chinese hamster lung cells [71]. Studies of OP and NP resulting in LOAELs and/or NOAELs are summarized in Table 8 and 9. Toxicity of OP in animal studies is probably due to estrogenic effects, resulting in changes in sperm morphology or testicular weight, earlier puberty or changed hormone levels [79, 80, 92]. Sharpe and co-workers reported decreases in average testis and prostate weight at the dose level 12.9 µg OP/kg bw/day in male Wistar rats following administration in drinking water, but they could not repeat the results in a later study [89, 132]. In another low-dose study, LOAEL for sperm tail abnormalities in male rats exposed to OP in drinking water, was 20 ng/kg bw/day [79]. This dose level is 1000-10 000 times lower than other reported low LOAELs (gestation length in pigs, onset of puberty in sheep), and no research group has tried to repeat the study. The critical effect of sperm tail abnormalities may not affect reproduction in rats, since no effect on reproduction capacity due to OP exposure has been reported in any long-term study in the rat. Another study performed according to US EPA guidelines did not reveal any effects on reproductive ability or sperm morphology in rats at dose levels up to 149 mg/kg bw/day [101]. If OP exposure may cause broken, coiled or bent sperm tails in man is unknown. In an OP study on female pigs, LOAEL for increased length of gestation was 0.01 mg/kg bw/day [88]. This study had few number of pigs in each exposure group due to a study design with few pigs in the first generation and crossing between different dose groups, making the interpretation of the results difficult. However, pigs have anatomical and physiological similarities with humans regarding metabolism and digestive system and effects of OP exposure in human may therefore be similar to those found in pigs. The critical effect (increased length of gestation) also considered of relevance for humans. Another study on lambs, have reported adverse effects at the dose levels 1.0 mg/kg bw/day when administered during fetal/postnatal life [93]. The effect of onset of puberty (measured as first progesterone rise in blood plasma followed by first estrus) could be a serious effect for women.

Table 8. Summary of toxicity studies of OP. s.c= subcutaneous administration. Species Route Duration Endpoint LOAEL NOAEL Ref. (days) mg/kg bw/day Short term studies SD rats s.c 9 Estrus cyclicity, 50 5.0 [94] Ovarian histology, Concentration LH in blood plasma Wistar rats2. water 21/42 Testis and ventral prostate weight 0.0129 0.00129 [89] Subchronic exposure Fischer rats water 120 Sperm tail abnormalities 0.00002 [79] Fischer rats s.c 30/60 LH, FSH, prolactin concentrations 341. [92] in blood plasma, Sperm count, abnormal sperms Sheep s.c 75/225/150 Onset of puberty 1.0 [93] Long term studies SD rats diet life time F0 and F1 bw 149 15 (200ppm) [101] SD rats diet life time Reproductive toxicity 149 [101] F0 Uterus weight Pigs s.c 63 Length of gestation 0.01 [88] 1. Exposure on body weight basis was calculated with the assumed body weight 250 g. 2. Results could not be repeated by same authors in a later study [90]. Although somewhat uncertain, the critical OP effect in males, sperm tail abnormalities, with a LOAEL of 20 ng/kg bw/day for rats will be used for the risk characterization for men. However, because of the uncertainty of human effects due to OP exposure and the lack of studies in low dose levels, more OP studies needs to be conducted. The critical OP effect in females, increased length of gestation, with a LOAEL of 10 000 ng/kg bw/day for pigs will be used for the risk characterisation for women. A total safety factor for OP is set 300 based on interspecies (×10) and intraspecies variation in toxicokinetics and toxicodymics (×10) and that LOAEL is used (×3).

daybwkgngdaybwkgngfactorSaftey

esabnormalittailspermratsLOAELmenTDI //067.0300

//20),()( ===

daybwkgngdaybwkgngfactorSaftey

lengthpregnancypigsLOAELwomenTDI //3.33300

//10000),()( ===

Studies on rats exposed to NP show that kidneys are target organs for both males and females, which may be affected at low doses (Table 9). The effect on kidney weight has been reported

37

38

reversible, after a recovery period of 4 weeks [80]. NP was most potent to rats exposed during the neonatal period, where a critical period has been suggested to the first two weeks of life [100]. Table 9. Summary of toxicity studies of NP. i.p= intraperitoneally administration Species Route Duration Endpoint LOAEL NOAEL Ref. (days) mg/kg bw/day Short term studies SD rats i.p 15 Testis, epididymis, 0.8 0.08 [100] seminal, prostate weight Subchronic exposure SD rats diet 90 bw, kidney weight 140 50 [80] Long term studies SD rats diet life time F0, F1, F2 kidney weight, 15 [102, 103] F2 sperm morphology SD rats diet life time Day of vaginal opening, 50 15 [102, 103] F1 uterine weight, F2 sperm density SD rats diet life time F0 bw, F1 polycystic kidney disease 61.51. 35.51. [104] SD rats diet life time F1 immunological changes 35-60 1.5-3.0 [105] CD-1 mice water life time F1 females kidney weight/bw, 50 µg/l [106] F1 males liver weight/bw 1. Lowest calculated mean dose during pregnancy, corresponding to 1000 and 500 ppm in diet, respectively. A good quality study of NP on rats has been performed in compliance with GLP and a study design according to OECD guidelines with the extension of a third generation [102, 103]. A LOAEL for increased kidney weight in both male and female rats was 200 ppm in diet (corresponding to about 15 mg/kg bw/day during non-reproductive phases) obtained in this study have been used in earlier risk assessments by the Danish Institute of Safety and Toxicology [127]. In the same study, NOAEL for reproductive effects (epididymal sperm density, abnormal sperm morphology, day of vaginal opening) was 15 mg/kg bw/day. Effects at lower doses have been reported in another study, administrating NP in drinking water to mice [106]. A NOAEL for decreased kidney weight (F1 females) and decreased liver weight (F1 males) was 50 µg/l (NP concentration in water). This study has a number of weaknesses, including few clear dose response effects, few individuals per group and there was no data on the quantity of water consumed. Another study using intraperitoneal administration to newborn rats, has reported reproductive effects at the dose level 0.08 mg/kg bw/day [100]. Because of very small groups no conclusion about NOAEL can be retrieved from this study. In the present study, the LOAEL for increased kidney weight in rats of 15 mg/kg bw/day will be used for further risk characterization. Since the TDI was calculated by the Danish Institute of Safety and Toxicology [127], new studies have been published: NP has been reported to be not genotoxic in several in vitro tests as well as in two in vitro studies, and is according to the European risk assessment report considered non mutagenic [1]. Two in vivo carcinogenic studies of NP on rats have been found [84, 85], indicating no carcinogenic effects at human exposure levels. Therefore the safety factor of 3000 used in the Danish report may be lowered.

A total safety factor for NP is set to 300 considering interspecies (×10) and intraspeices variation in toxickinetics and toxicodynamics (×10) and because a LOAEL is used (×3).

daybwkgngdaybwkgmgfactorSaftey

weightkidneyincreasedratLOAELTDI //00050300

//15),(===

Discussion Swedish perch contains 6-7 times higher concentration of NP than of OP. This is similar to reported AP concentrations in fish elsewhere [28, 33, 42, 49], where NP generally is present in concentrations of 10-fold or higher than OP. Detectable concentrations of OP were found in most freshwater fish but not in the marine species herring and flounder (detection limit 0.3 ng/g f.wt), which suggest a higher contamination of APs in freshwater fish than in marine fish. High concentrations of OP and NP are measured in areas with high pollution due to release of wastewater from households (and/or industries) [40, 45, 48]. In Sweden, pooled samples of bream and perch from waters influenced by effluents from wastewater treatment plants had median OP and NP concentration of 10.5 and 15 ng/g f.wt, respectively (see Table 4). Comparing average OP levels in marine fish, Swedish flounder (0.15 ng/g f.wt) has 10-fold lower OP concentration than Italian common sole (1.2-1.7 ng/g f.wt) and about the same difference is found between Swedish herring (0.15 ng/g f.wt) and anchovy from the Italian coast (0.3-1.7 ng/g f.wt) [42], suggesting higher AP contamination in the Mediterranean Sea than in Swedish marine waters. Mean concentrations of OP in Swedish bream from Viskan River is about 11 times higher compared to OP concentrations in bream from German rivers influenced by wastewater effluent (up to 0.65 ng/g f.wt) 2001 and about 2 times higher than the highest measured bream concentration found bream from German rivers 1994 (5.5 ng/g f.wt) [49]. Generally, NP concentrations in Swedish perch are low compared to earlier reported concentrations in freshwater fish. In carp from the Cuyahoga River in Ohio (US), concentrations of NP from a clean site were about 5 times higher (6.6 ng/g f.wt) than the mean concentration in Swedish perch (1.34 ng/g f.wt) [45]. From the same river, downstream release of effluent from a wastewater treatment plant, concentrations of NP in carp were 6-7 times higher (100 ng/g f.wt) than in Swedish perch from the contaminated Lake Ringsjön (15 ng/g f.wt). NP levels of 15 ng/g f.wt in perch from Ringsjön in Sweden is somewhat higher than concentrations up to 9.8 ng/g f.wt in bream from German rivers influenced by sewage water effluents 2001, but about 7 times lower than the highest measured concentration (112 ng/g f.wt) in bream from the German River Saar 1994 [49]. NP concentration in perch from Lake Ringsjön is about twice as high as reported NP concentrations in fish from Kalamazoo River in Michigan (US), receiving secondary and tertiary wastewater treatment effluents and industrial discharges (including those from paper manufacturing) [44].

39

In scenario 1, average intake of OP from freshwater fish for men and women was low, since most individuals did not eat freshwater fish at all (Table 6). Compared to the TDI of 0.067 ng/kg bw/day for men suggested in present study based on the LOAEL for sperm tail abnormalities in rats (20 ng/kg bw) [79] and a safety factor 300, only one man was just above tolerable intake (0.0787 ng/kg bw/day). No woman was at risk of exceeding the TDI of 33.3 ng/kg bw/day suggested in present study based on a LOAEL in pigs (length of gestation) of 10 000 ng/kg

40

bw/day [88] and a safety factor 300. The MOS between LOAEL in rats (sperm tail abnormalities) and the highest individual exposure for men was 250. MOS for women was above 2.28×105 compared to a LOAEL in pigs (length of gestation), or 2.28×107 compared to LOAEL in sheep (onset of puberty). In scenario 2, average intake of OP from fish species with known concentration was up to 1.23 ng per person and day, or 0.0186 ng/kg bw/day (Table 6 and 7). One woman exceeded the earlier suggested TDI of 0.5 ng/kg bw/day [13] and two men exceeded the TDI suggested in present study of 0.067 ng/kg bw/day. No woman exceeded the TDI of 33.3 ng/kg bw/day suggested in present study. MOS between highest individual exposure among men and LOAEL in rats (sperm tail abnormalities) is 220-fold. MOS for women was above 28 600 compared to a LOAEL in pigs (length of gestation), or 100-fold higher compared to a LOAEL in sheep (onset of puberty). In scenario 3, the estimation of OP intake from all fish in Sweden of 6.11-7.51 ng per person and day is about 10 times lower compared to a calculated intake of OP from seafood for the Italian population (50 ng/day) [42]. Compared the TDI of 0.067 ng/kg bw/day in present study, about half of the men exceeded the tolerable intake (Table 7). None of the women exceeded the TDI of 33.3 ng/kg bw/day, where the maximum exposure for one woman (1.36 ng/kg bw/day) is about 30 times lower than the tolerable intake. A MOS for men between LOAEL in rats (sperm tail abnormalities) and the highest individual exposure (1.36 ng/kg bw/day) was about 15, which suggest a low margin of safety. Compared to LOAEL in pigs (length of gestation) MOS for women was high ~9200 and even higher compared LOAEL in sheep (onset of puberty) ~917 000. Average worst-case intake of OP from fish in scenario 4 (206-235 ng/day) was about 4-5 times higher than the calculated daily intake from seafood in Italy (50 ng/day) [42]. According to the TDI of 0.067 ng/kg bw/day all men that eat heavily contaminated fish will exceed the tolerable intake. One woman of age 41-75 year had an OP exposure above the TDI of 33.3 ng/kg bw/day suggested in present study. A MOS between the highest OP exposure among men and LOAEL in rats (sperm tail abnormalities) is about 1. For women, MOS between the highest individual exposure (45 ng/kg bw/day) and LOAEL in pigs (length of gestation) was about 200, about 22 000 compared to LOAEL in sheep (onset of puberty). In scenario 5, the average intake of NP from freshwater fish only was low (0.00858-0.0129 ng/kg bw/day), because most individuals did not eat freshwater fish. Compared to the TDI of 50 000 ng/kg bw/day suggested in present study based on a LOAEL for increased kidney weight in rats of 15 mg/kg bw/day [103] and a safety factor 300, no risk for human effects is suggested. In scenario 6, worst-case intake of NP from Swedish fish (294-336 ng/day) was 40-fold lower than a calculated daily intake from seafood for the Italian population (12 200 ng/day) [42], and about 25 times lower than an estimated intake of NP from overall food for the German population (7500 ng/day) [8]. Comparing worst-case intake of NP from fish for 95 % of Swedish adults (about 13 ng/kg bw/day) with the TDI of 50 000 ng/kg bw/day suggested in present study, the intake was about 1 100 times lower. MOS between LOAEL for increased kidney weight and the highest individual exposure (64.9 ng/kg bw/day) was about 230 000, suggesting no risk for adverse human effects.

41

Generally, women were more exposed to OP and NP from fish than men. However, the difference was small, and the average intake of APs for women in scenario 4 and 6 was only about 1-1.5 ng/kg bw/day more than men. In scenario 3, the highest contribution to human exposure was from farmed salmon, when using measured OP concentration of artic char in calculations. A rough estimate of the total OP intake from food and water may be calculated. In a German food study [8], about 3.15 % of total intake of NP (7.5 µg/ day) was from fish/fish products. Assuming that OP intake from fish in scenario 3 (mean intake about 0.1 ng/kg bw/day) corresponds to 3.15 % of the total intake from food, the OP intake from overall food would be 3.2 ng/kg bw/day. Using OP concentrations in Swedish drinking water, the intake from water (0.571 ng/kg bw/day, see section human intake/ biomonitoring) and from overall food is 3.75 ng/kg bw/day. This intake exceeds the TDI for men of 0.067 ng/kg bw/day, but is about 8-9 times lower than the TDI for women of 33.3 ng/kg bw/day. A MOS for men compared to the LOAEL in rats (sperm tail abnormalities) is about 6.3. For women, MOS compared to LOAEL in pigs (length of gestation) is about 3200-fold. As the fish consumption in Sweden is higher than that in Germany and therefore probably has a higher % contribution to the total OP intake from food, this calculation is most likely an overestimation. As estimation of total NP exposure for non-occupational humans, an exposure from overall food, drinking water and migration from food rapping plastics may be calculated: A German NP intake from overall food was estimated to 7.5 µg/day [8]. Assuming similar food consumption in Sweden, the NP intake corresponds to about 110 ng/kg bw/day for a 70 kg adult. Further, based on measured NP concentration in Swedish drinking water, the daily intake from water is calculated to 8.29 ng/kg bw/day (see section human intake/ biomonitoring), and the intake from food wrapping plastics has been estimated to 200 ng/kg bw/day [1]. From above data, the total NP intake is calculated to 315 ng/kg bw/day. Compared to the TDI of 50 000, the intake is about 160 times lower, suggesting no risk for non-occupationally exposed humans. In the present study, molluscs and crustaceans were not included in the intake calculations. According to Italian studies, molluscs may have higher concentrations of OP and NP than fish [41, 42]. In a rough estimate of daily intake of NP from molluscs for the general Italian population (assumed consumption 20-40 g/day), NP intake was 6000-13000 ng/day, about same level as the daily intake from overall seafood (12 200 µg/day), estimated in a later study by the same authors. However, a low consumption of shellfish in Sweden (about 3-5 g/day) [131], and concentrations of APs in common mussels from the Baltic sea of 0.3 ng/g f.wt for OP and below detection limit 2.0 ng/g f.wt for NP (2001) [48], suggest a low exposure from shellfish for the Sweden population. There was an obvious lack of fish data, which made estimation of AP intake in different scenarios somewhat uncertain. The consumption statistic of fish may contain sources of errors, especially regarding the origin of salmon consumed. The estimated OP intake in scenario 3 (using median concentration of artic char for farmed, self-caught and Baltic salmon) is probably an overestimation, since no detectable OP concentration in rainbow trout (farmed in Swedish waters) was found. If instead half of the detection limit for OP in rainbow tout (0.06 ng/g f.wt) is used in calculations, mean total intake is about 2-3 times lower. In that case, about 95% of the men would be below the tolerable intake, but individual men with high fish consumption would

42

still exceed the TDI of 0.067 ng/kg bw/day. For women, the MOS would be about 15 000 compared to LOAEL in pigs (length of gestation). As a conclusion, the daily intake of OP from fish for the Swedish population is probably in the range of 0.809-7.51 ng per person and day, or 0.0186-0.115 ng/kg bw/day (scenario 2 and 3). Intake of OP from fish is high compared with suggested TDI for men, but the animal data behind the proposed TDI are uncertain. The single OP study of effects on sperm quality in rats needs to be repeated before firm conclusions can be drawn about sperm effects at low OP exposure in animals. In rat studies, no decreased reproductive ability due to OP exposure has been reported. Whether OP exposure causes negative sperm effects in man is unknown. The risk for serious effects on women, such as shortened length of gestation or earlier puberty is low for OP intake from fish only. Considering combined exposure of OP from overall food and drinking water, there is a low safety margin between exposure levels causing sperm effects in rats and intake levels in humans, but a high safety margin for effects on length of gestation for women. For NP, worst-case intake from fish is up to 779 ng per person and day or 12.7 ng/kg bw/day, which is low in comparison to the proposed TDI. The risk for adverse effects in humans due to exposure from fish intake is low with a MOS of about 230 000. The risk from combined exposure from overall food, drinking water and food wrapping plastic is low for non-occupational humans.

43

References 1. ECB, European union risk assessment report, 4-nonylphenol (branched) and

nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

2. Ahel, M., W. Giger, and M. Koch, Behavior of Alkylphenol Polyethoxylate Surfactants in the Aquatic Environment 1. Occurrence and Transformation in Sewage-Treatment. Water Research, 1994. 28(5): p. 1131-1142.

3. Jungclaus, G.A., V. Lopez-Avila, and R.A. Hites, Organic compounds in an industrial wastewater: A case study of their environmental impact. Environmental science & technology, 1978. 12(1).

4. Staples, C.A., et al., Evaluation of aquatic toxicity and bioaccumulation of C8- and C9-alkylphenol etoxylates. Environmental toxicology and chemistry, 1998. 17: p. 2470-2480.

5. White, R., et al., Environmentally persistent alkylphenolic compounds are estrogenic. Endocrinology, 1994. 135(1): p. 175-82.

6. Gronen, S., et al., Serum vitellogenin levels and reproductive impairment of male Japanese Medaka (Oryzias latipes) exposed to 4-tert-octylphenol. Environ Health Perspect, 1999. 107(5): p. 385-90.

7. Bergstedt, O., N. Paxéus, and H. Rydberg, Miljöföroreningar i dricksvatten- Inledande studie exemplifierad med alkylfenoler, ftalater och PAH, VAV AB, Editor. 1999.

8. Guenther, K., et al., Endocrine disrupting nonylphenols are ubiquitous in food. Environmental Science & Technology, 2002. 36(8): p. 1676-1680.

9. Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

10. Darnerud, P.O. and J. Hajslova, Alkylfenoler i reningsverksprover och fisk, in Nationellt miljögiftsseminarium. 2002: Nässjö.

11. Oxford university, Safety (MSDS) data for 4-tert-butylphenol, The physical and theoretical chemistry laboratory Oxford university, Editor. 2005.

12. OECD SIDS, SIDS initial assessment report for 10th SIAM, P-tert-butylphenol, CAS N: 98-54-4, UNEP Publications, Editor. 2000.

13. Baser, M., et al., Risk assessment on 4-tert-octylphenol. 2004, Toxikologiprogrammet, Karolinska institutet: Solna.

14. Kemikalieinspektionen, SPIN-database, The Swedish National Chemicals Inspectorate, Editor. 2005.

15. Kemikalieinspektionen, Hazard assessments- Chemical substances selected in the Swedish Sunset project, The Swedish National Chemicals Inspectorate, Editor. 1995.

16. SIDS, SIDS initial assessment report for 3rd SIAM, Phenol, 4-(1,1,3,3,tetramethylbutyl), SIDS, Editor. 1995. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

17. Rosqvist, L., Supervisors: Andreasson F. Länsstyrelsen in Skåne, and Christensen B. Umeå universitet, Screening av fenoler i Skånes miljö, Utvärdering av provtagning 2003 i reningsverk, sjöar och hav, Länsstyrelsen i Skåne län, Editor. 2003.

18. SNV-rapport 3492, Identifiering av miljöstörande ämnen i avfall, Naturvårdsverket, Editor. 1988.

19. European union, EG-directive 76/769/EEG. 2003.

44

20. Remberger, M., et al., Screening tertiary butylphenols, menthylphenols and long-chain alkylphenols in the Swedish environment, IVL, Editor. 2003.

21. Ferreira-Leach, A.M. and E.M. Hill, Bioconcentration and distribution of 4-tert-octylphenol residues in tissues of the rainbow trout (Oncorhynchus mykiss). Mar Environ Res, 2001. 51(1): p. 75-89.

22. Liber, K., M.L. Knuth, and F.S. Stay, An integrated evaluation of the persistence and effects of 4-nonylphenol in an experimental littoral ecosystem. Environmental Toxicology and Chemistry, 1999. 18(3): p. 357-362.

23. Eklund, R., et al., Biodegradation of 4-nonylphenol in seawater and sediment. Environ. Pollut., 1993. 79: p. 59-61. In ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

24. Giger, W., P.H. Brunner, and C. Schaffner, 4-Nonylphenol in sewage sludge: accumulation of toxic metabolites from nonionic surfactants. Science, 1984. 225(4662): p. 623-5. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

25. Chang, B.V., F. Chiang, and S.Y. Yuan, Biodegradation of nonylphenol in sewage sludge. Chemosphere, 2005. 60(11): p. 1652-9.

26. Crespin, M.A., M. Gallego, and M. Valcarcel, Continuous microwave-assisted extraction, solvent changeover and preconcentration of monophenols in agricultural soils. Journal of Chromatography A, 2000. 897(1-2): p. 279-293.

27. Dachs, J., D.A. Van Ry, and S.J. Eisenreich, Occurrence of estrogenic nonylphenols in the urban and coastal atmosphere of the lower Hudson River estuary. Environmental Science & Technology, 1999. 33(15): p. 2676-2679.

28. Basheer, C., H.K. Lee, and K.S. Tan, Endocrine disrupting alkylphenols and bisphenol-A in coastal waters and supermarket seafood from Singapore. Mar Pollut Bull, 2004. 48(11-12): p. 1161-7.

29. Sheldon, L.S. and R.A. Hites, Organic compounds in the Delawere river. Environmental science and technology, 1978. 12: p. 1188-1194. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

30. Blackburn, M.A. and M.J. Wardock, Concentrations of alkylphenols in rivers and esttuaries in England and Wales. Water research, 1995. 29(7): p. 1623-1629. In. ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

31. Naturvårdsverket, Höga halter av miljöfarliga ämnen i miljön? Resultat från miljöövervakningens screeningprogram 1996-2003. 2005.

32. Fooken, C., M. Häckl, and P. Seel, Orientierende Messungen gefährlicher Stoffe. Landesweite Untersuchungen auf organische Spurenverunreinigungen in hessischen Flissgewässern, Abwässern und Klärschlämmen, 1991-1995, H.l.f. umwelt, Editor. 1995. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

45

33. Lye, C.M., et al., Estrogenic alkylphenols in fish tissues, sediments and waters from the U.K. Tyne and Tees estauries. Environmental Science & Technology, 1999. 33: p. 1009-1014.

34. Bund-/Länderausschluss für die Berwertug von Umweltchemikalien (BLAU): Bericht der arbeitsgruppe `Datensammlung für die Bewertung von Umweltchemikalien´. 1995. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

35. Van Ry, D.A., et al., Atmospheric seasonal trends and environmental rate of alkylphenols in the lower Hudson River Estuary. Environmental Science & Technology, 2000. 34(12): p. 2410-2417.

36. McLeese, D.W., et al., Lethality and accumulation of alkylphenols in aquatic fauna. Chemosphere, 1981. 10: p. 723-730.

37. Freitag, D., et al., Ecotoxicological profile analysis. VII. Screening chemicals for their environmental behavior by comparative evaluation. Ecotoxicology and Environmental Saftey, 1982. 6(1): p. 60-81. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

38. Brooke, L.T., Accumulation and lethality for two freshwater fishes (fatheaded minnow and bluegill) to nonylphenol, USEPA draft report EPA contract No 68-CI-0034, Editor. 1993. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

39. Eklund, R., et al., Bioaccumulation of 4-nonylphenol in marine animals- A re-evaluation. Environmental Pollution, 1990. 64: p. 107-120.

40. Ahel, M., J. McEvoy, and W. Giger, Bioaccumulation of the Lipophilic Metabolites of Nonionic Surfactants in Fresh-Water Organisms. Environmental Pollution, 1993. 79(3): p. 243-248.

41. Ferrara, F., et al., Alkylphenolic compounds in edible molluscs of the Adriatic Sea (Italy). Environ Sci Technol, 2001. 35(15): p. 3109-12.

42. Ferrara, F., et al., Alkylphenols and alkylphenol ethoxylates contamination of crustaceans and fishes from the Adriatic Sea (Italy). Chemosphere, 2005. 59(8): p. 1145-1150.

43. Yang, D.K. and W.H. Ding, Determination of alkylphenolic residues in fresh fruit and vegetables by extractive steam distillation and gas chromatography-mass spectrometry. Journal of Chromatography A, 2004. 1088: p. 200-204.

44. Keith, T.L., et al., Identification and quantitation of nonylphenol ethoxylates and nonylphenol in fish tissues from Michigan. Environ Sci Technol, 2001. 35(1): p. 10-3.

45. Rice, C.P., et al., Alkylphenol and alkylphenol-ethoxylates in carp, water and sediment from the Cyahoga River, Ohio. Environ. Sci. Technol., 2003. 37: p. 3747-3754.

46. Turrini, A., et al., Food consumption patterns in Italy: the INN-CA Study 1994-1996. Eur J Clin Nutr, 2001. 55(7): p. 571-88. In: Ferrara, F., et al., Alkylphenols and alkylphenol ethoxylates contamination of crustaceans and fishes from the Adriatic Sea (Italy). Chemosphere, 2005. 59(8): p. 1145-1150.

47. Ferrara, F. and E. Funari, Rapporti ISTISAN. 1999 In: Ferrara, F., et al., Alkylphenolic compounds in edible molluscs of the Adriatic Sea (Italy). Environ Sci Technol, 2001. 35(15): p. 3109-12.

46

48. Wenzel, A., et al., Restrospective monotoring of alkyl monoethoxylates in aquatic biota from 1985-2001: Results from the German environmental speicmen bank, supporting information to manuscript. Enviromental science and technology, 2004.

49. Wenzel, A., et al., Retrospective monitoring of alkylphenols and alkylphenol monoethoxylates in aquatic biota from 1985 to 2001: Results from the German Environmental Specimen Bank. Environmental Science & Technology, 2004. 38(6): p. 1654-1661.

50. Kirchmann, H. and A. Tengsved, Organic Pollutants in Sewage-Sludge .2. Analysis of Barley Grains Grown on Sludge-Fertilized Soil. Swedish Journal of Agricultural Research, 1991. 21(3): p. 115-119. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

51. Gilbert J. et al, Identification by gas chromatography-mass spectrometry of vinyl chloride oligomers and other low-molecular-weight components in poly(vinyl chloride) resins for food packing applications. Journal of Chromatography, 1982. 237: p. 249-261. In: Bergstedt, O., N. Paxéus, and H. Rydberg, Miljöföroreningar i dricksvatten- Inledande studie exemplifierad med alkylfenoler, ftalater och PAH, VAV AB, Editor. 1999.

52. Gilbert J. et al, Compositional analysis of commersial PVC bottles and studies of aspects of specific and overall migration into foods and stimulants. Food additives and contaminants, 1986. 3(2): p. 133-144. In: Bergstedt, O., N. Paxéus, and H. Rydberg, Miljöföroreningar i dricksvatten- Inledande studie exemplifierad med alkylfenoler, ftalater och PAH, VAV AB, Editor. 1999

53. Muller, S., P. Schmid, and C. Schlatter, Pharmacokinetic behavior of 4-nonylphenol in humans. Environmental Toxicology and Pharmacology, 1998. 5(4): p. 257-265.

54. Kosaka, M., et al., Urinary Metabolite Levels in Workers Handling P-Tert-Butylphenol as an Index of Personal Exposure. International Archives of Occupational and Environmental Health, 1989. 61(7): p. 451-455. In: OECD SIDS, SIDS initial assessment report for 10th SIAM, P-tert-butylphenol, CAS N: 98-54-4, UNEP Publications, Editor. 2000.

55. Yang, F. and R.E. Boissy, Effects of 4-tertiary buthylphenol on the tyrosinase activity in human melanocytes. Pigment cell research, 1999. 12: p. 237-245. In: OECD SIDS, SIDS initial assessment report for 10th SIAM, P-tert-butylphenol, CAS N: 98-54-4, UNEP Publications, Editor. 2000.

56. Certa, H., et al., Toxicokinetics of p-tert-octylphenol in male Wistar rats. Arch Toxicol, 1996. 71(1-2): p. 112-22.

57. Upmeier, A., et al., Toxicokinetics of p-tert-octylphenol in female DA/Han rats after single i.v. and oral application. Arch Toxicol, 1999. 73(4-5): p. 217-22.

58. Monteiro-Riviere, N., et al., Comparative in vitro dermal absorption of nonylphenol and nonylphenol etoxylates (NPE4 and NPE9) trough human skin, porcine and rodent skin. Submitted for publication. 1999. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

59. Pedersen, R.T. and E.M. Hill, Identification of novel metabolites of the xenoestrogen 4-tert-octylphenol in primary rat hepatocytes. Chem Biol Interact, 2000. 128(3): p. 189-209.

47

60. Klaassen, C.D., Casarett and Doull´s Toxicology: The basic science of poisons. 6th ed, ed. C.D. Klaassen. 2001: McGraw-Hill.

61. Koster, H., et al., Dose-dependent shifts in the sulfation and glucuronidation of phenolic compounds in the rat in vivo and in isolated hepatocytes. The role of saturation of phenolsulfotransferase. Biochem Pharmacol, 1981. 30(18): p. 2569-75. In: OECD SIDS, SIDS initial assessment report for 10th SIAM, P-tert-butylphenol, CAS N: 98-54-4, UNEP Publications, Editor. 2000.

62. Daidoji, T., et al., Glucuronidation and excretion of nonylphenol in perfused rat liver. Drug Metab Dispos, 2003. 31(8): p. 993-8.

63. Knaak, J.B., J.M. Eldridge, and L.J. Sullivan, Excretion of certain polyethylene glycol ether adducts of nonylphenol by the rat. Toxicol Appl Pharmacol, 1966. 9(2): p. 331-40. In: SNV-rapport 3492, Identifiering av miljöstörande ämnen i avfall, Naturvårdsverket, Editor. 1988.

64. Doerge, D.R., et al., Mass spectrometric determination of p-nonylphenol metabolism and disposition following oral administration to Sprague-Dawley rats. Reproductive Toxicology, 2002. 16(1): p. 45-56.

65. Birgersson, B., O. Stemer, and E. Zimerson, Kemiska hälsorisker Toxikologi i kemiskt perspektiv. 2nd ed. 1995, Malmö: Liber ekonomi.

66. Klonne, D.R., et al., Acute Toxicity and Primary Irritation of Para-Tertiary Butylphenol. Drug and Chemical Toxicology, 1988. 11(1): p. 43-54.

67. IUCLID, Existing chemicals substance, ID. 140-66-9, European commission. European chemicals bureau, Editor. 2000.

68. Gaworski, C.L., E.R. Kimbead, and R.L. Doyle, Acute toxicity of a number of a number of chemicals of interest to the Air Force, University of California extension Wright Patterson Air Force base, Editor. 1979. In: OECD SIDS, SIDS initial assessment report for 10th SIAM, P-tert-butylphenol, CAS N: 98-54-4, UNEP Publications, Editor. 2000.

69. Smyth, H.F., Jr. , et al., Range-finding toxicity data: List VII. Am Ind Hyg Assoc J, 1969.

30(5): p. 470-6. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

70. Dean, B.J., et al., Genetic toxicology testing of 41 industrial chemicals. Mutat Res, 1985. 153(1-2): p. 57-77. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

71. Ministry of Health and Welfare: Japan, Toxicity testing reports of environmental chemicals 4. 1996. p. 227-304. In: OECD SIDS, SIDS initial assessment report for 10th SIAM, P-tert-butylphenol, CAS N: 98-54-4, UNEP Publications, Editor. 2000.

72. Tuschl, H., et al., Investigations into the inhibition of DNA repair process by detergents. Environmental Physiology and Biochemistry, 1975. 5: p. 84-91. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

73. Garberg, P., E.L. Akerblom, and G. Bolcsfoldi, Evaluation of a genotoxicity test measuring DNA-strand breaks in mouse lymphoma cells by alkaline unwinding and hydroxyapatite elution. Mutat Res, 1988. 203(3): p. 155-76. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

48

74. Matsuoka, A., T. Sofuni, and M. Ishidate, Effects on surfactants on the induction of chromosomal aberrations in chinese hamster cells in culture. Mutation Research, 1986. 164: p. 273-274. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

75. Buttar, H.S., S.H.H. Swierenga, and T.I. Matula, Evaluation of the Cytotoxicity and Genotoxicity of the Spermicides Nonoxynol-9 and Octoxynol-9. Toxicology Letters, 1986. 31(1): p. 65-73. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

76. Hüls, A., In vitro mammal cell gene mutation test with nonylphenol. 1990. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

77. Hüls, A., In vivo mouse micronucleus test. 1999. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

78. Hüls, A., Mutagenitätsuntersuchung von Nonylphenol im Mikrokern-Test. 1988. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

79. Blake, C.A., et al., Effects of 4-tert-octylphenol given in drinking water for 4 months on the male reproductive system of Fischer 344 rats. Reprod Toxicol, 2004. 18(1): p. 43-51.

80. Cunny, H.C., et al., Subchronic toxicity (90-day) study with para-nonylphenol in rats. Regulatory Toxicology and Pharmacology, 1997. 26(2): p. 172-178.

81. Hirose, M., et al., Modification of N-methyl-N'-nitro-N-nitrosoguanidine-induced forestomach and glandular stomach carcinogenesis by phenolic antioxidants in rats. Cancer Res, 1988. 48(18): p. 5310-5. In: OECD SIDS, SIDS initial assessment report for 10th SIAM, P-tert-butylphenol, CAS N: 98-54-4, UNEP Publications, Editor. 2000.

82. Yoshida, M., et al., Induction of different types of uterine adenocarcinomas in Donryu rats due to neonatal exposure to high-dose p-t-octylphenol for different periods. Carcinogenesis, 2002. 23(10): p. 1745-50.

83. Kohno, H., et al., Lack of modifying effects of 4-tert-octylphenol and benzyl butyl phthalate on 3,2 '-dimethyl-4-aminobiphenyl-induced prostate carcinogenesis in rats. Cancer Science, 2004. 95(4): p. 300-305.

84. Inaguma, S., et al., p-nonylphenol pretreatment during the late neonatal period has no effect on 3,2 '-dimethyl-4-aminobiphenyl-induced prostate carcinogenesis in male F344 rats. Cancer Letters, 2004. 212(2): p. 159-166.

85. Fukamachi, K., et al., Possible enhancing effects of atrazine and nonylphenol on 7,12-dimethylbenz[a]anthracene-induced mammary tumor development in human c-Ha-ras proto-oncogene transgenic rats. Cancer Sci, 2004. 95(5): p. 404-10.

86. Soto, A.M., et al., an ´in culture´ bioassay to assess the estrogenicity of xenobiotics (E-screen). 1992. In: White, R., et al., Environmentally persistent alkylphenolic compounds are estrogenic. Endocrinology, 1994. 135(1): p. 175-82.

49

87. Matthews, J., et al., Differential estrogen receptor binding of estrogenic substances: a species comparison. Journal of Steriod Biochemistry & Molecular Biology, 2000. 74: p. 223-234.

88. Bogh, I.B., et al., Endocrine disrupting compounds: Effect of octylphenol on reproduction over three generations. Theriogenology, 2001. 55(1): p. 131-150.

89. Sharpe, R.M., et al., Gestational and lactational exposure of rats to xenoestrogens results in reduced testicular size and sperm production. Environ Health Perspect, 1995. 103(12): p. 1136-43.

90. Sharpe, R.M., K.J. Turner, and J.P. Sumpter, Endocrine disruptors and testis development. Environ Health Perspect, 1998. 106(5): p. A220-1.

91. Katsuda, S., et al., Irreversible effects of neonatal exposure to p-tert-octylphenol on the reproductive tract in female rats. Toxicology and Applied Pharmacology, 2000. 165(3): p. 217-226.

92. Blake, C.A. and F.R. Boockfor, Chronic administration of the environmental pollutant 4-t-octylphenol to adult male rats interferes with the secretion of luteinizing hormone, follicle stimulating hormone, prolactin and testosterone. Biology of Reproduction, 1997. 57: p. 255-266.

93. Wright, C., et al., Effect of maternal exposure to the environmental estrogen, octylphenol during fetal and/or postnatal life on onset of puberty, endocrine status, and ovarian follicular dynamics in Ewe lambs. Biology of Reproduction, 2002. 67: p. 1734-1740.

94. Willoughby, K.N., et al., Neonatally administered tert-octylphenol affects onset of puberty and reproductive development in female rats. Endocrine, 2005. 26(2): p. 161-8.

95. Odum, J., et al., Effects of p-nonylphenol (NP) and diethylstilboestrol (DES) on the Alderley Park (Alpk) rat: comparison of mammary gland and uterus sensitivity following oral gavage or implanted mini-pumps. J Appl Toxicol, 1999. 19(5): p. 367-78. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

96. Odum, J., et al., Comparative activities of p-nonylphenol and diethylstilbestrol in noble

rat mammary gland and uterotrophic assays. Regul Toxicol Pharmacol, 1999. 29(2 Pt 1): p. 184-95.

97. Colerangle, J.B. and D. Roy, Exposure of environmental estrogenic compound nonylphenol to noble rats alters cell-cycle kinetics in the mammary gland. Endocrine, 1996. 4(2): p. 115-122. In: Öberg, M., et al., Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

98. de Jager, C., M.S. Bornman, and G. van der Horst, I. The effect of p-nonylphenol, an environmental toxicant with oestrogenic properties, on fertility potential in adult male rats. Andrologia, 1999. 31(2): p. 99-106. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

99. de Jager, C., M.S. Bornman, and J.M.C. Oosthuizen, II. The effect of p-nonylphenol on the fertility potential of male rats after gestational, lactational and direct exposure. Andrologia, 1999. 31(2): p. 107-113. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European

50

Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

100. Lee, P.C., Disruption of male reproductive tract development by administration of the xenoestrogen, nonylphenol, to male newborn rats. Endocrine, 1998. 9(1): p. 105-11.

101. Tyl, R.W., et al., Two-generation reproduction study with para-tert-octylphenol in rats. Regulatory Toxicology and Pharmacology, 1999. 30(2): p. 81-95.

102. Chapin, R.E., et al., The effects of 4-nonylphenol in rats: a multigeneration reproduction study. Toxicol Sci, 1999. 52(1): p. 80-91.

103. NTP, Nonylphenol: multigenerational reproductive effects in Sprague-Dawley rats when exposed to nonylphenol in the diet. 1997.

104. Latendresse, J.R., et al., Polycystic kidney disease induced in F-1 Sprague-Dawley rats fed para-nonylphenol in a soy-free, casein-containing diet. Toxicological Sciences, 2001. 62(1): p. 140-147.

105. Karrow, N.A., et al., Nonylphenol alters the activity of splenic NK cells and the numbers of leukocyte subpopulations in Sprague-Dawley rats: a two-generation feeding study. Toxicology, 2004. 196(3): p. 237-245.

106. Kyselova, V., et al., Effects of p-nonylphenol and resveratrol on body and organ weight and in vivo fertility of outbred CD-1 mice. Reprod Biol Endocrinol, 2003. 1: p. 30.

107. Brooke, L.T., Acute and chronic toxicity of nonylphenol to ten species of aquatic organisms, USEPA draft report, Editor. 1993, EPA. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

108. Kopf, W., Wirkung endokriner stoffe in biotests mit wasserorganismen. In Stoffe mit endokrinerwirkung in wasser, Institut für Wasserforschung München, Editor. 1997: Oldenbourg. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

109. IUCLD, ECB- Existing chemicals, IUCLID data base, Editor. 1995. In: Öberg, M., et al.,

Toxicological evaluation of octylphenol and butylphenol (draft version). 2005, Institute of Environmental Medicine, Karolinska Institutet.

110. Weinberger, P. and M. Rea (1981), Nonylphenol: A pertubant additive to an aquatic ecosystem. From proceedings of the 7th aquatic workshop, Montreal, 1981. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

111. McLeese, D.W., Zitko V., and M. Peterson, Structure-lethality relationships for phenols, anilines and other aromatic compounds in shrimp and clams. Chemosphere, 1979. 9: p. 53-57.

112. Holcombe, G.W., et al., The acute toxicity of selected substituted phenols, benzenes and benzoic acid esters to fatheaded minnows Phimphales promelas. Environmental Pollution (Series A), 1984. 35: p. 367-381. In: ECB, European union risk assessment report, 4-nonylphenol (branched) and nonylphenol, in 2nd priority list, volume 10, European Commission Joint Research Centre, Editor. 2002, Institute for health and consumer protection.

51

113. Kelly, S.A. and R.T. Di Giulio, Developmental toxicity of estrogenic alkylphenols in killifish (Fundulus heteroclitus). Environmental Toxicology and Chemistry, 2000. 19(10): p. 2564-2570.

114. Jobling, S., et al., Inhibition of testicular growth in rainbow trout (Oncorhynchus mykiss) exposed to estrogenic alkylphenolic chemicals. Environmental Toxicology and Chemistry, 1996. 15(2): p. 194-202.

115. Madsen, L.L., B. Korsgaard, and P. Bjerregaard, Estrogenic effects in flounder Platichthys flesus orally exposed to 4-tert-octylphenol. Aquat Toxicol, 2003. 64(4): p. 393-405.

116. Rasmussen, T.H. and B. Korsgaard, Estrogenic octylphenol affects seminal fluid production and its biochemical composition of eelpout (Zoarces viviparus). Comp Biochem Physiol C Toxicol Pharmacol, 2004. 139(1-3): p. 1-10.

117. Pickford, K.A., et al., Route of exposure affects the oestrogenic response of fish to 4-tert-nonylphenol. Aquat Toxicol, 2003. 65(3): p. 267-79.

118. Harvey, P.W., A. Cockburn, and K.C. Rush, Endocrine and hormonal toxicology. 1999, Chichester: Wiley. xxvii,568.

119. Chumakov, N.N., G.P. Babanov, and A.G. Smirnov, [On vitiligo-like dermatoses in workers in the phenol-formaldehyde resin works.]. Vestn Dermatol Venerol, 1962. 36: p. 3-8. In: OECD SIDS, SIDS initial assessment report for 10th SIAM, P-tert-butylphenol, CAS N: 98-54-4, UNEP Publications, Editor. 2000.

120. Forck, G., et al, Arch. Derm. Res., 1981. 270: p. 224. In: OECD SIDS, SIDS initial assessment report for 10th SIAM, P-tert-butylphenol, CAS N: 98-54-4, UNEP Publications, Editor. 2000.

121. Ebner, H., et al., [Vitiligo from p-tert. butylphenol; a contribution to the problem of the internal manifestations of this occupational disease]. Derm Beruf Umwelt, 1979. 27(4): p. 99-104. In: OECD SIDS, SIDS initial assessment report for 10th SIAM, P-tert-butylphenol, CAS N: 98-54-4, UNEP Publications, Editor. 2000.

122. Gebhart, W., T. Luger, and G. Niebauer, [Vitiligo due to p-tert-butylphenol]. Ann Dermatol Venereol, 1980. 107(8-9): p. 809-14. In: OECD SIDS, SIDS initial assessment report for 10th SIAM, P-tert-butylphenol, CAS N: 98-54-4, UNEP Publications, Editor. 2000.

123. Bendsen, E., et al., Effect of 4-octylphenol on germ cell number in cultured human fetal gonads. Human Reproduction, 2001. 16(2): p. 236-243.

124. Muller, S., P. Schmid, and C. Schlatter, Evaluation of the estrogenic potency of nonylphenol in non-occupationally exposed humans. Environmental Toxicology and Pharmacology, 1998. 6(1): p. 27-33.

125. Clark, L.B., et al., Determination of alkylphenol ethoxylates and their acetic acid derivates in drinking water by particle beam liquid chromatography/mass spectrometry. Int. J. Environ. Anal. Chem., 1993. 47: p. 167-180. In: Muller, S., P. Schmid, and C. Schlatter, Evaluation of the estrogenic potency of nonylphenol in non-occupationally exposed humans. Environmental Toxicology and Pharmacology, 1998. 6(1): p. 27-33.

126. Petrick, G., D.E. Schulzbull, and J.C. Duinker, Analyses of Polychlorinated-Biphenyls and Organic Contaminants in Drinking-Water. Zeitschrift Fur Wasser- Und Abwasser-Forschung-Journal for Water and Wastewater Research-Acta Hydrochimica Et Hydrobiologica, 1992(2): p. 115-121. In: Muller, S., P. Schmid, and C. Schlatter, Evaluation of the estrogenic potency of nonylphenol in non-occupationally exposed humans. Environmental Toxicology and Pharmacology, 1998. 6(1): p. 27-33.

52

127. Nielsen, E., et al., Toxicological evaluation and limit values for nonylphenol, nonylphenol ethoxylates, tricresyl, phosphates and benzoic acid. 2000, The institute of food Safety and Toxicology, Danish Veterinary and Food administration.

128. Hajslova, J., Determination of alkylphenols and musk compounds in waste water, sediments and fish from Sweden. 2002, Institute of chemical technology, department of food chemistry and analysis: Prague.

129. Darnerud, P.O., Resultatrapport till Miljöövervakningsenheten vid Naturvårdsverket: Screening av alkylfenoler, särskilt oktylfenol, i prover från reningsverk och i fiskprover, Livsmedelsverket, Editor. 2002.

130. Homepage of project Ringsjön, http://www.ringsjon.se/ringfakta.asp, Projekt Ringsjön, Ringsjöfakta, Hörby kommun, Höörs kommun, Eslövs kommun, Lunds universitet, Länsstyrelsen i Skåne län, Sydvatten, Editor. 2005.

131. Becker, W. and M. Pearson, Riksmaten 1997-98, Kostvanor och näringsintag i Sverige, metod och resultatanalys,, Livsmedelsverket, Editor. 2002, Avdelningen för information och nutrition Livsmedelsverket.

132. Rudel, R.A., et al., Identification of alkylphenols and other estrogenic phenolic compounds in wastewater, septage, and groundwater on Cape Cod, Massachusetts. Environmental Science & Technology, 1998. 32(7): p. 861-869.